# Ask AI
Source: https://docs.sweetspotgov.com/ai/ask-ai

Chat with any solicitation, pursuit, or your whole organization. Answers come with citations.

<img alt="Ask AI in Sweetspot" />

Ask AI is Sweetspot's assistant. It reads solicitations, searches your Library and pursuit files, queries federal spending data, and browses the web, then answers with citations to its sources.

## Where you can chat

Ask AI runs in three contexts, and the context determines what the AI can see:

| Context          | What the AI works with                                                      |
| ---------------- | --------------------------------------------------------------------------- |
| **Organization** | Your Library, company profile, and all of Sweetspot's data sources          |
| **Opportunity**  | A specific solicitation and its documents, plus organization context        |
| **Pursuit**      | Everything in the pursuit workspace: files, notes, requirements, and drafts |

Start an organization chat from **Ask AI** in the sidebar, or open a chat from any opportunity or pursuit.

<Note>
  Chats are stored where you start them. A chat started inside a pursuit has all the context from that pursuit's files, and it lives in the pursuit's Ask AI tab: it won't appear in the sidebar's Ask AI chat list.
</Note>

## The chat workspace

The chat is a multi-pane workspace: your conversation, the files in scope, and the sources behind each answer sit side by side. You can keep multiple chat tabs open, expand AI-generated tables (headers stay pinned while you scroll), and copy results out.

## Agentic behavior

* **Subagents.** For questions that require digging, the AI runs subagents that search your pursuit and files for requirements, compliance details, and supporting documents, then report back with citations linked to the source.
* **Interactive clarifications.** When your request is ambiguous, the AI presents clickable options instead of guessing.
* **Tool use.** The AI chooses among Library search, pursuit search, federal award data, connected integrations, and web search based on your question. You can also direct it explicitly; see the [Prompting Guide](/prompting-guide).

## Collaborate in chat

Share a chat with teammates as view-only or as collaborators: multiple people and the AI in one thread, in real time. New people default to view only; switch them to **Can collaborate** to let them participate. You can also copy the full chat to your clipboard.

<img alt="The Share Chat dialog with view-only and collaborate permissions" />

<Note>
  Share responsibly: responses may draw on files some teammates can't access, and follow-up questions may reveal details beyond the current transcript.
</Note>

## Reuse what works

Save your best instructions as [Playbooks](/ai/playbooks) and apply them in any chat with one click.

<Tip>
  Answer quality depends on what's in your [Organization Library](/library/organization-library). Fit and gap analysis in particular relies on your uploaded past performance and capability documents.
</Tip>


# Playbooks
Source: https://docs.sweetspotgov.com/ai/playbooks

Save your best AI prompts and reuse them anywhere

<img alt="Playbooks in Sweetspot" />

A Playbook is a saved, named AI instruction. When a prompt works well, save it as a Playbook and reuse it instead of retyping it. Playbooks can be shared across your organization.

## Create a Playbook

<img alt="Creating a new Playbook: write it yourself or generate with AI" />

<Steps>
  <Step title="Create a new Playbook">
    Go to **Library → Playbooks** and click **New Playbook**.
  </Step>

  <Step title="Write it yourself or generate it with AI">
    Either write the instruction directly, or describe in natural language what you want the Playbook to do and Sweetspot AI writes it for you. Good Playbooks are specific about the task, the format of the output, and the sources the AI should use.
  </Step>

  <Step title="Apply it in any chat">
    Quick-apply any Playbook from the chat window with one click, in any context: organization, opportunity, or pursuit.
  </Step>
</Steps>

## Playbook ideas

* **Bid/no-bid first pass**: "Summarize this opportunity's scope, ceiling, set-aside, key dates, and evaluation criteria. Then assess our fit against our company profile and flag any gaps."
* **Amendment review**: "List every change in the latest amendment and its impact on our draft response."
* **Past performance pull**: "Search within my organization for past performance relevant to this scope and summarize the three strongest matches."
* **Section L/M check**: "Extract all Section L submission instructions and Section M evaluation factors into a table."

## Best practices

Treat a Playbook as something you iterate on, not something you write once:

1. **Write it with Ask AI.** Describe the outcome you want and let the AI draft the Playbook.
2. **Run it on a real opportunity or pursuit.** Test against the kind of work you'll actually use it for.
3. **Review the output.** Note what's missing, extra, or in the wrong format.
4. **Describe the changes to the AI.** For example: "always produce 5 win themes for me to reference when running the win theme development Playbook."
5. **Repeat.** A few rounds of this gets you a Playbook that produces the output you're looking for every time.

<Tip>
  Shared Playbooks are a way to standardize your team's process: everyone runs the same qualification, amendment review, and compliance prompts.
</Tip>


# Bookmarks
Source: https://docs.sweetspotgov.com/discover/bookmarks

Save interesting opportunities to a shared watch list

Bookmarks let you save an opportunity without adding it to your pipeline. Use them for opportunities you want to keep visible while you decide whether to pursue them.

## How Bookmarks work

* **Save with one click.** Bookmark any opportunity from search results or its detail page.
* **Shared with your team.** Bookmarks appear in a dashboard visible to your whole organization, which avoids two people triaging the same opportunity.
* **Convert when ready.** When you decide to bid, convert the bookmark into a pursuit with one click. The opportunity moves into your [Pursuits pipeline](/pursuits/dashboard) with its documents attached.

## When to bookmark vs. pursue

| Situation                                               | Action                                                            |
| ------------------------------------------------------- | ----------------------------------------------------------------- |
| Looks interesting, needs triage                         | Bookmark                                                          |
| Waiting on more information (amendment, Q\&A responses) | Bookmark, and add a [Monitor](/intelligence/monitors) for updates |
| You've decided to bid, or need to track capture work    | Add as a pursuit                                                  |

This split keeps pipeline metrics accurate: pursuits contain only opportunities your team is actively working, and bookmarks hold everything else.


# Saved Searches
Source: https://docs.sweetspotgov.com/discover/saved-searches

Automate your best queries and get alerted when new opportunities match

<img alt="Saved Searches in Sweetspot" />

A Saved Search runs your query on a schedule and alerts you when new opportunities match.

## Create a Saved Search

<Steps>
  <Step title="Create the query">
    Choose the opportunity type (Federal Contracts, Federal Grants, DIBBS, SBIR/STTRs, or State & Local), enter your keyword query, and add filters like NAICS codes, set-asides, or agencies.
  </Step>

  <Step title="Set the alert options">
    * **Email interval**: how often Sweetspot emails you new matches.
    * **Minimum opportunity relevancy**: how closely a match must fit before it's included. High is the typical choice; use medium for a broader list.
    * **Email recipients**: who on your team receives the alerts.
  </Step>

  <Step title="Save it">
    Save the search. It's named after the query itself, so a descriptive query doubles as a recognizable name in your Saved Searches list.
  </Step>

  <Step title="Review new matches">
    Go through the notifications in your email or in Sweetspot's [Inbox](/intelligence/inbox) as they arrive.
  </Step>
</Steps>

Manage all of your Saved Searches from the **Saved Searches** page in the Search flyout.

## Tips

* **Create one search per need.** Narrow queries produce fewer, more relevant alerts than one broad query.
* **Filter by your set-asides.** If you hold certifications like 8(a), HUBZone, WOSB, or SDVOSB, add them as filters so alerts only include opportunities you qualify for.
* **Delete searches you no longer need.** Stale searches add alert noise.

## Beyond Saved Searches

Saved Searches watch for **new opportunities matching a query**. To watch specific vendors, agencies, live opportunities, or web sources, use [Monitors](/intelligence/monitors). Monitor alerts and Saved Search results both land in your [Inbox](/intelligence/inbox).


# Search
Source: https://docs.sweetspotgov.com/discover/search

Natural-language opportunity search across federal, DIBBS, SBIR, grant, and state and local sources

<img alt="Opportunity search in Sweetspot" />

Search accepts plain-language queries and returns opportunities from SAM.gov, Grants.gov, DIBBS, and over 1,000 state and local sources. It matches on meaning, not just keywords.

```
Find IT modernization contracts with the Navy over $5M
```

## Opportunity types

The Search flyout in the sidebar covers five opportunity sources, each with its own search page:

| Type                        | What it covers                                                                                            |
| --------------------------- | --------------------------------------------------------------------------------------------------------- |
| **Federal Contracts**       | Contract opportunities from SAM.gov: solicitations, presolicitations, sources sought, and special notices |
| **Federal Grants**          | Grant opportunities from Grants.gov                                                                       |
| **DIBBS Contracts**         | DLA DIBBS solicitations, searchable by NSN and part number                                                |
| **SBIR / STTRs**            | Small Business Innovation Research and Technology Transfer topics                                         |
| **State & Local Contracts** | SLED opportunities aggregated from 1,000+ state and local procurement portals                             |

<Tip>
  Set your default opportunity type under **Settings → Preferences** so the Search icon always lands where you work most.
</Tip>

## Searching effectively

* **Write descriptive queries.** "Cloud migration services for civilian agencies" returns better results than a list of disconnected keywords.
* **Use filters for structured criteria.** NAICS and PSC codes, set-asides, buyers, due dates, and solicitation numbers all live in the filter panel. For a pure code or number lookup, set the filter and leave the query blank.
* **Check the amendment history.** When a solicitation is amended, Sweetspot highlights the differences between versions so you can see exactly what changed.

See the [Prompting Guide](/prompting-guide) for more search tips.

## From result to pursuit

From any opportunity you can:

* **Open the full details**: response deadline, set-aside, notice type, AI summary, solicitation details and documents, point of contact, project timeline, and linked awards showing the current recipient.
* **Ask AI** about it: requirements, evaluation criteria, incumbent history, and fit, with cited answers. See [Ask AI](/ai/ask-ai).
* **Generate a capture brief** for a structured bid/no-bid analysis with win probability and gap assessment.
* **Bookmark it** to a shared watch list. See [Bookmarks](/discover/bookmarks).
* **Add it as a pursuit** to start working it in your pipeline. See [Pursuits](/pursuits/dashboard).

## Research the competitive landscape

To research who currently holds the work, including incumbents, award history, expiring contracts, and agency buying patterns, use [Federal Market Intelligence](/intelligence/federal-market-intelligence).


# Federal Market Intelligence
Source: https://docs.sweetspotgov.com/intelligence/federal-market-intelligence

Search 167M+ federal awards to research incumbents, recompetes, forecasts, and agencies

<img alt="An AI-built dashboard of DHS cybersecurity recompetes in Sweetspot" />

Federal Market Intelligence is Sweetspot's research layer over federal spending data from USASpending, FPDS, and SAM. It covers more than 167 million awards. Use it to find who holds the work today, what they were paid, when it recompetes, and how an agency buys.

## Federal Award AI

Ask spending questions in plain English and get answers with charts, insights, and exportable results:

```
Show me SDVOSB IT contracts over $5M awarded by the VA in the last two years
```

The AI translates your question into a query over the awards dataset and returns results you can drill into and export.

The same questions work in [Ask AI](/ai/ask-ai), if chat is your preferred way to work. Ask there when you want to combine award data with other context, like your Library or an opportunity's documents, in one conversation; use Federal Award AI when you want a dedicated research surface with charts and exports.

## Recipients

Research any contractor or grantee. Search by name, UEI, DUNS, or CAGE code, and filter by socioeconomic certifications (8(a), HUBZone, WOSB, SDVOSB), NAICS, PSC, agency, location, and spending range.

<img alt="A recipient profile in Sweetspot with insights and award cash flow" />

Each recipient profile shows total obligations, cash-flow trends, agency mix, set-aside participation, past awards, and parent/subsidiary hierarchies. Export result sets to Excel or CSV.

Use Recipients to research incumbents before a bid, find teaming partners with verified past performance, and track competitors.

## Award Details

Every contract and grant has a detail page: transaction history, obligation timeline, period of performance, and sub-awards. Linked solicitations connect historical awards to live opportunities, which surfaces recompetes before a new solicitation posts.

<img alt="An award detail page in Sweetspot with funding progress and key dates" />

## Forecasts

Agency procurement forecasts list planned purchases, often 12 or more months before an RFP posts. Filter by agency, NAICS, and set-aside. Each forecast has a detail page with the timeline, estimated dates, value, and points of contact.

<img alt="A forecast detail page in Sweetspot with timeline, dates, and contacts" />

## Budgets

Search federal budget accounts to see where funding is growing or shrinking before it's obligated. Each account shows the budget lifecycle (requested, enacted, appropriated, obligated, outlaid), how much of it is addressed to vendors, and year-over-year trends.

<img alt="Federal budget account search in Sweetspot" />

## Protests

Search GAO bid protest decisions. Each protest shows the protester, agency, solicitation, filing and decision dates, and the outcome (denied, dismissed, or sustained).

<img alt="GAO bid protest search in Sweetspot" />

## Contract Vehicles

Research GWACs, IDIQs, and Schedules. Each vehicle page shows total obligated, ceiling, and remaining amounts, the full awardee list, task orders, and analytics on obligations, market share, agencies, set-asides, competition, and geography.

<img alt="A contract vehicle page in Sweetspot with awardees and obligation analytics" />

## NAICS

Every NAICS code has its own workspace showing market size, year-over-year momentum, and vendor count, with tabs for top buyers, incumbents, recompetes, open opportunities, and peer codes. The Recompetes tab lists expiring contracts with obligated amounts, remaining ceiling, and days left, so you can find recompete targets in your industry.

<img alt="A NAICS code workspace in Sweetspot showing expiring contracts" />

## PSC

PSC codes get the same workspace, opening with a market brief: market size over the last 12 months, momentum, incumbent concentration, and how much of the market is set aside for small business, plus obligations by fiscal year and which agencies are driving the change.

<img alt="A PSC code workspace in Sweetspot with market brief and money flow" />

## Agency dashboards

Every agency has a dashboard with fiscal-year obligations, contract actions, small business set-aside percentage, and grant awards, plus a monthly spending trend, award type mix, top recipients, and NAICS and PSC breakdowns.

<img alt="The NASA agency dashboard in Sweetspot" />


# Inbox
Source: https://docs.sweetspotgov.com/intelligence/inbox

Where monitor alerts and incoming opportunities arrive

The Inbox collects monitor alerts, incoming GSA eBuy RFQs, and Saved Search matches in one feed. It sits at the top of the sidebar, with an unread indicator when something new arrives.

## Sources

* **Monitors.** Every alert from your [Monitors](/intelligence/monitors): new awards, amendments, agency activity, and web mentions.
* **Saved Searches.** New opportunities matching your [Saved Searches](/discover/saved-searches).
* **GSA eBuy.** If your organization has the eBuy integration enabled, RFQ notifications forwarded to your unique Sweetspot ingest address appear here automatically with their attachments. See [Integrations](/settings/integrations).

## Triage

Tools for working through the feed:

* Filter by monitor, alert type, and status
* Full-text search across alerts
* Star, archive, mark read/unread, and dismiss, individually or in bulk
* A detail view per alert with a lineage timeline showing the history of changes that led to it

From any alert you can jump straight to the underlying opportunity, award, or vendor, ask [Ask AI](/ai/ask-ai) to qualify it, or add it to your [Pursuits](/pursuits/dashboard).

<Tip>
  If you keep dismissing the same kind of alert, tighten that monitor's filters instead. See [Monitors](/intelligence/monitors).
</Tip>


# Monitors
Source: https://docs.sweetspotgov.com/intelligence/monitors

Watch vendors, agencies, opportunities, and the open web. Updates come to you.

<img alt="Monitors in Sweetspot" />

A Monitor watches a vendor, agency, opportunity, or web source and sends new awards, amendments, and updates to your [Inbox](/intelligence/inbox) and notification channels. Set one up once and Sweetspot checks for changes continuously.

Create monitors from the **Monitors** page, from any vendor, agency, or opportunity page, or let the AI build them for you (see [Set up with AI](#set-up-with-ai)).

## Choose what to track

Creating a monitor starts with picking one of four types:

<CardGroup>
  <Card title="Vendor & Competitor Monitors" icon="building">
    Watch a company, like a competitor or teaming partner. You're alerted when it wins a new prime award, IDV, or OTA, when its SAM registration changes, or when it's mentioned on the web.
  </Card>

  <Card title="Opportunity Monitors" icon="file-contract">
    Watch a specific opportunity. You're alerted when an amendment posts, Q\&A responses are released, the deadline changes, or new attachments are added. Works on contract opportunities, grants, and forecasts.
  </Card>

  <Card title="Agency & Office Monitors" icon="landmark">
    Watch an agency or office you sell to. You're alerted when it posts a new solicitation, forecast entry, award, or grant notice.
  </Card>

  <Card title="Web & Keyword Monitors" icon="globe">
    Watch the open web for a topic, described in plain English, like "zero trust modernization from DHS, DoD, and GSA". Sweetspot checks daily and alerts you on new mentions. Also useful for watching state procurement sites for SLED RFPs.
  </Card>
</CardGroup>

Then choose the specific thing to watch. Vendor and agency monitors take a company or an agency; web monitors take a plain-English description of the topic.

For example, opportunity monitors also ask which kind of record you're tracking:

* **Contract**: a prime award
* **IDV**: an IDIQ or contract vehicle
* **OTA / OT IDV**: an Other Transaction award or vehicle
* **Grant**: a grant opportunity

Each option lists the events it watches, like modifications, ceiling changes, amendments, and deadline moves.

<img alt="Choosing the type of record an opportunity monitor tracks" />

## Coverage: what triggers an alert

Once you've chosen what to track, you configure the monitor's coverage: the set of **signals** that trigger alerts. Each signal has a toggle. Switch one off to stop alerting on it; at least one signal must stay on.

<img alt="Monitor coverage signals with per-signal filter rows" />

Below each signal is an optional **Filters** row:

* **No filters** (the default): the signal alerts on every update.
* **Add filter**: narrows that one signal without affecting the others. Available filters depend on the signal and include keywords, NAICS and PSC codes, set-asides, notice types, award and IDV types, places of performance, and a minimum obligated amount.

## Subscribers: who gets alerted

Finally, name the monitor and choose where its alerts go:

* **Email**: select the teammates who get an email when this monitor fires.
* **Slack**: pick a connected Slack channel to post alerts to. Requires the [Slack integration](/settings/integrations).
* **Teams**: coming soon.

<img alt="Choosing monitor subscribers across email, Slack, and Teams" />

Every alert also lands in your [Inbox](/intelligence/inbox), regardless of subscriber settings.

## Set up with AI

Instead of configuring monitors one at a time, use **Set up my monitors with AI**. The AI asks about your business: what you sell, which agencies you target, who your competitors and teaming partners are, and where you're the incumbent. Then it recommends and creates monitors for you.


# Introduction
Source: https://docs.sweetspotgov.com/introduction

The Sweetspot user guide

<img alt="Sweetspot" />

<img alt="Sweetspot" />

Sweetspot is an AI platform for government contractors. It brings opportunity discovery, federal market intelligence, pipeline management, and proposal writing into one secure workspace at [console.sweetspotgov.com](https://console.sweetspotgov.com).

This guide covers every part of the console, organized the way the product is: **Search**, **Ask AI**, **Pursuits**, **Library**, and your **Inbox**.

## Set up

Start by giving Sweetspot context about your business. Search results, capture briefs, and drafts all use your Library and company profile as input.

<CardGroup>
  <Card title="Quickstart" icon="rocket" href="/quickstart">
    Set up your account, run your first search, and qualify an opportunity.
  </Card>

  <Card title="Organization Library" icon="folder-open" href="/library/organization-library">
    Upload past proposals, capability statements, and resumes so the AI can work from your institutional knowledge.
  </Card>
</CardGroup>

## Find and qualify opportunities

<CardGroup>
  <Card title="Search" icon="magnifying-glass" href="/discover/search">
    Natural-language search across federal contracts, grants, DIBBS, SBIR/STTRs, and state and local sources.
  </Card>

  <Card title="Saved Searches" icon="bell" href="/discover/saved-searches">
    Automate your best queries and get alerted when new matches post.
  </Card>

  <Card title="Federal Market Intelligence" icon="chart-line" href="/intelligence/federal-market-intelligence">
    Research 167M+ federal awards, incumbents, forecasts, budgets, protests, and contract vehicles.
  </Card>

  <Card title="Monitors" icon="radar" href="/intelligence/monitors">
    Watch vendors, agencies, opportunities, and the open web. Updates come to your Inbox.
  </Card>
</CardGroup>

## Work your pipeline

<CardGroup>
  <Card title="Ask AI" icon="sparkles" href="/ai/ask-ai">
    Chat with any solicitation, pursuit, or your whole organization. Answers come with citations.
  </Card>

  <Card title="Pursuits" icon="list-check" href="/pursuits/dashboard">
    Track every opportunity from discovery to award with dashboard, list, board, and calendar views.
  </Card>

  <Card title="Proposal Engine" icon="file-lines" href="/proposals/proposal-engine">
    Shred the RFP, build compliance and capability matrices, and draft in a Word-like editor.
  </Card>

  <Card title="Integrations" icon="plug" href="/settings/integrations">
    Connect Microsoft 365, Salesforce, GovWin, GSA eBuy, and Slack.
  </Card>
</CardGroup>

## Security

Sweetspot is SOC 2 Type II audited and CMMC 2.0 Level 2 certified, with FedRAMP Moderate Authorization in progress. AI requests are served from U.S. data centers with zero data retention, and your data never trains third-party models. Role-based access control and SSO are available for enterprise organizations.

For additional security information, request access to our trust center at [trust.sweetspot.so](https://trust.sweetspot.so/).

## Getting help

Start with this guide: the answer to most how-do-I questions is on the page covering that feature, and the search bar above works across every page. If you're still stuck, click **Live Support** in the console to chat directly with the Sweetspot team. Questions go straight to real people, and response times are fast.


# Example Library Setup
Source: https://docs.sweetspotgov.com/library/folder-structure

Any folder structure works: what matters is specific subfolders the AI can search precisely

How you organize the Library matters more than which structure you pick. Any folder structure works well, as long as each folder is specific enough that you (and the AI) can point at exactly the right one when searching.

## The rule: subfolders over mega-folders

* **Keep folders under roughly 20 files.** When one grows past that, split it into subfolders.
* **Folders within folders is good.** A mega-folder with 100+ files is not.
* **Why:** when a search is scoped to a folder (see the [Prompting Guide](/prompting-guide)), a small, specific folder means the AI reads only the relevant files, so answers come back faster, more accurate, and easier to verify.

<Tree>
  <Row icon="❌" name="Avoid: one mega-folder" />

  <Row icon="📂" name="02_Past-Performance" />
</Tree>

<Tree>
  <Row icon="✅" name="Better: folders within folders, under 20 files each" />

  <Row icon="📂" name="02_Past-Performance" />

  <Row icon="📁" name="DHS_Border-IT_70CMSW21C0000123_2019-2024" />

  <Row icon="📁" name="01_Official-Record" />

  <Row icon="📁" name="02_PP-Writeups" />

  <Row icon="📁" name="03_Performance-Evidence" />

  <Row icon="📁" name="VA_Claims-Modernization_36C10B22C0000045_2021-2025" />

  <Row icon="📁" name="01_Official-Record" />

  <Row icon="📁" name="02_PP-Writeups" />

  <Row icon="📁" name="03_Performance-Evidence" />
</Tree>

## A proven example

The rest of this page is the structure Sweetspot recommends as a starting point: five top-level folders, each with a single job. Adopt it as is, or keep your own layout and apply the same rule.

<Tree>
  <Row icon="📂" name="01_Company-Identity" />

  <Row icon="📂" name="02_Past-Performance" />

  <Row icon="📂" name="03_Proposal-Library" />

  <Row icon="📂" name="04_People" />

  <Row icon="📂" name="05_Teaming-Partners" />
</Tree>

## 01\_Company-Identity

Who you are: the facts and approved language every proposal draws from.

<Tree>
  <Row icon="📂" name="01_Company-Identity" />

  <Row icon="📁" name="Corporate-Facts" />

  <Row icon="📁" name="Capability-Statements" />

  <Row icon="📁" name="Approved-Boilerplate" />
</Tree>

## 02\_Past-Performance

Your contract execution record only: proof of what you've done. One folder per contract, named `{"{Agency}_{Program}_{Contract#}_{YYYY-YYYY}"}`.

<Tree>
  <Row icon="📂" name="02_Past-Performance" />

  <Row icon="📁" name="DHS_Border-IT-Modernization_70CMSW21C0000123_2019-2024" />

  <Row icon="📁" name="01_Official-Record" />

  <Row icon="📁" name="02_PP-Writeups" />

  <Row icon="📁" name="03_Performance-Evidence" />
</Tree>

## 03\_Proposal-Library

What you wrote: submitted proposals only, never edited after close-out, plus a curated reference folder of excerpts worth reusing.

Name each proposal folder `{"{YYYY}_{Agency}_{Sol#}_{ShortTitle}"}` plus the outcome: `WIN`, `LOSS`, `NO-AWARD`, or `PENDING`.

<Tree>
  <Row icon="📂" name="03_Proposal-Library" />

  <Row icon="📁" name="_Reference-Content" />

  <Row icon="📁" name="Management-Approach" />

  <Row icon="📁" name="Staffing-and-Recruiting" />

  <Row icon="📁" name="Transition-Plans" />

  <Row icon="📁" name="Technical-Cybersecurity" />

  <Row icon="📁" name="Graphics-and-Exhibits" />

  <Row icon="📁" name="2024_GSA_47QTCA24R0001_Cloud-Migration_WIN" />

  <Row icon="📁" name="01_Solicitation" />

  <Row icon="📁" name="02_Final-Submission" />

  <Row icon="📁" name="03_Debrief" />
</Tree>

## 04\_People

The team: one trusted source for resumes and credentials.

<Tree>
  <Row icon="📂" name="04_People" />

  <Row icon="📁" name="Resume-Masters" />

  <Row icon="📁" name="Bios-and-Certs" />

  <Row icon="📁" name="Org-Charts-LCATs" />
</Tree>

## 05\_Teaming-Partners

Partner materials only: never mixed into Past Performance, Proposal Library, or People. One folder per partner.

<Tree>
  <Row icon="📂" name="05_Teaming-Partners" />

  <Row icon="📁" name="Redstone-Analytics" />

  <Row icon="📁" name="01_Corporate-Info" />

  <Row icon="📁" name="02_Their-Past-Performance" />

  <Row icon="📁" name="03_Their-Resumes" />

  <Row icon="📁" name="_Agreements" />
</Tree>

## Why the example works

* **One job per folder.** When each folder holds one kind of content, the AI (and your teammates) always know where to look, and where new files belong.
* **Names carry the facts.** Contract numbers, date ranges, and win/loss outcomes in folder names make search results self-explanatory and let you scope AI requests precisely: "search the 02\_Past-Performance folder for cloud migration work."
* **Submitted work stays frozen.** Separating the never-edited record (`03_Proposal-Library`) from reusable excerpts (`_Reference-Content`) means you can reuse confidently without corrupting the archive.
* **No folder gets huge.** Every branch bottoms out in small, specific folders, which is the rule that makes everything else work.


# Memories
Source: https://docs.sweetspotgov.com/library/memories

Durable preferences and conventions the AI recalls in every future session

Memories are durable preferences, conventions, and facts that Sweetspot AI recalls in future sessions, so you stop repeating yourself. Tell it once that summaries stay under one page, or that proposals use active voice and your approved customer terminology, and every future chat and draft respects it.

Find them under **Library → AI Context → Memories**.

## What to save as a memory

Save guidance that will still be true weeks from now:

* **Instructions**: "always do X", "never do Y", tone and format rules, corrections you keep making
* **Conventions**: terminology, naming, templates, review practices
* **Stable facts**: about your organization, your role, and how your team works
* **Preferences**: working-style preferences that apply broadly

Don't save one-off task status or anything tied to a single conversation; that's what the chat itself is for.

## How memories get created

* **You describe it.** From the Memories page, describe what Sweetspot should remember. The AI saves it in chat, and if an existing memory already covers it, it updates that one instead of creating a duplicate. Editing works the same way: describe the change and Sweetspot applies it in chat.
* **The AI suggests them.** As you work, Sweetspot proposes memories it thinks are worth keeping (marked **Agent**, versus **Manual** for ones you created). Suggestions wait in a review queue; approve the good ones, reject the rest. Admins can restrict suggestion approval to admins only.

## Scopes

Every memory belongs to a scope, which controls where it applies:

| Scope            | Applies to                                                        |
| ---------------- | ----------------------------------------------------------------- |
| **Organization** | Every chat and draft across your org                              |
| **Pursuit**      | Work inside one pursuit                                           |
| **File**         | One specific file; shows up as a note when the AI reads that file |

## The toggles

* **Personal vs. Shared.** A personal memory applies only to your sessions. Share it in its scope to apply it to the whole team; shared memories require a stability note explaining why the guidance holds over time (for example, "stable team voice").
* **Standing.** Standing guidance is included in every AI session in its scope, always. Non-standing memories are recalled only when relevant. Reserve standing for the rules that must never be missed, like proposal voice.
* **Enabled.** Disable a memory to switch it off without deleting it; archive it to remove it entirely.

## Import from another AI assistant

Already trained ChatGPT or Claude on your preferences? The **import card** on the Memories page gives you a prompt to copy into the other assistant; paste its export back into Sweetspot and the AI saves the durable entries, skips the stale ones, and summarizes what it kept.


# Organization Library
Source: https://docs.sweetspotgov.com/library/organization-library

Your organization's documents and reusable content, indexed for AI search and drafting

<img alt="Organization Library in Sweetspot" />

The Library holds your organization's documents inside Sweetspot: past proposals, capability statements, resumes, and reusable content. Everything in it is indexed, and the AI uses it in searches, capture briefs, and drafts.

The Library sidebar is organized into four groups: **Organization Data**, **AI Context**, **Proposal Templates**, and **Directory**.

## Organization Data

### Company Profile

A profile of your organization: what you do, your differentiators, certifications, and past performance themes. Generate it from your uploaded documents, then edit it. The AI includes it as context in every conversation.

### Files

The document repository, and the foundation everything else builds on:

* **Upload** by drag and drop, and organize with folders and tags. Create subfolders to help the AI narrow down when searching: "search the Past Performance folder" beats searching everything. Any structure works if you keep folders small and specific; see the [Example Library Setup](/library/folder-structure).
* **What to include**: past proposals, capability statements, key personnel resumes, past performance writeups and CPARS, pricing sheets, and approved boilerplate.
* **Files feed the AI.** Capture briefs, capability scoring, and proposal drafts all pull from what's here: the more complete your Files, the better every output.

### SharePoint

If your documents live in SharePoint, connect the [Microsoft 365 integration](/settings/integrations) and browse your SharePoint files directly from **Library → SharePoint** instead of re-uploading them.

## AI Context

### Playbooks

Saved AI instructions, shareable across the team, applied in any chat with one click. See [Playbooks](/ai/playbooks).

### Memories

Durable preferences and conventions the AI recalls in every future session: proposal voice, formatting rules, approved terminology, and stable facts about your organization. See [Memories](/library/memories).

## Proposal Templates

* **Excerpts**: reusable blocks of text for proposals: boilerplate, corporate overviews, win themes, and management approaches you use again and again.
* **Word Templates**: branded .docx style templates applied when exporting to Word: your fonts, headers, footers, and table of contents.

## Directory

### Companies

Your directory of primes, subs, teaming partners, and competitors, with UEI and CAGE identifiers, relationship types, and interaction history. Link companies to the pursuits they're involved in. See [Companies & Contacts](/pursuits/companies-and-contacts) for the full CRM.

### People

One trusted source for the people behind your proposals and relationships:

* **Two kinds of records**: internal team members and external contacts, each with name, title, and contact details, plus preferred names and aliases.
* **Resumes and credentials**: attach an existing resume or upload a new one; track bios, certifications, and clearance status. When drafting, the AI matches personnel to solicitation requirements from these records.
* **Relationship context**: associate a person with a company or agency, and track prioritization, recent outreach, and tags for external contacts.
* **Bulk import and cleanup**: import people from an Excel file, and merge duplicates when they appear.

## Keeping it current

* **Set up the right folders first.** Small, specific folders beat mega-folders; see the [Example Library Setup](/library/folder-structure) for the rule and a proven example layout.
* **Start with the essentials.** Your last 5 to 10 winning proposals, current capability statement, and key personnel resumes cover most drafting needs.
* **Keep the latest version in the Library.** When a capability statement, resume, or rate card changes, make sure the Library copy is the current one, so the AI is always drafting from today's content, not last year's.
* **Add submitted proposals back in.** After each submission, add the final volumes to the Library so the next pursuit can draw on your most recent work, wins and losses alike.


# Prompting Guide
Source: https://docs.sweetspotgov.com/prompting-guide

Get the best results from Sweetspot's AI search, chat, and proposal writing

Sweetspot AI is agentic: it chooses tools, runs searches, and reads documents on your behalf. Clear, detailed prompts get better results because they tell the AI which tools to use and what to look for.

<Tip>
  Found a prompt that works well? Save it as a [Playbook](/ai/playbooks) so you and your team can reuse it with one click.
</Tip>

## Searching

* **Avoid acronyms.** The search can often expand them, but spelling things out is more reliable.
* **Detail is critical.** "Cybersecurity services" returns broad results; "zero trust network modernization for DoD" returns useful ones.
* **Use filters for structured criteria.** Solicitation numbers, NAICS codes, PSC codes, and buyer names belong in the filter panel, not the search bar. For those lookups, set the filters and leave the query blank.

## Chatting with Ask AI

* The AI picks the best tools for your question based on context. To force a search, say so explicitly: "search for ...".
* Specificity matters. "Section L instructions" produces better outcomes than "section L".
* Use the AI's subagents to your advantage: ask it to dig through your pursuit documents or your organization's files for compliance details, requirements, and past performance. Answers come back with citations to the source.
* When the AI asks a clarifying question with clickable options, picking one narrows the work and improves the answer.

## Focusing the AI's search

Tell the AI where to look:

* Within your organization's Library: `Search within my organization for ...`
* Within the current pursuit: `Search within my pursuit for ...`
* Both at once: `Search within the pursuit and organization for ...`
* The open web: `Use the internet to look up ...`

You can combine sources in one request, for example: `Use my pursuit files and also web search to compare the incumbent's approach with current CISA guidance.`

## Point at specific folders and files

The narrower the scope, the faster and more accurate the AI is. Searching your entire Library means examining every indexed document; naming a folder, tag, or file means the AI only reads what's relevant:

* A specific folder: `Search the "Past Performance" folder in my library for cloud migration projects`
* A specific file: `Use our 2025 capability statement to answer this`
* A tag: `Look only at files tagged "DHS" for relevant past performance`

This matters most in large Libraries. "Search my organization" makes the AI sift through every document; "search the Resumes folder" goes straight to the right set.

Clear folders and consistent tags make this kind of scoping possible; see [Organization Library](/library/organization-library).


# Proposal Engine
Source: https://docs.sweetspotgov.com/proposals/proposal-engine

Outline, draft, and edit your proposal with AI grounded in your Library

The Proposal Engine takes a solicitation from upload to draft: Sweetspot builds a compliance-mapped outline from the RFP, writes the draft section by section from your organization's Library, and hands it off to a Word-like editor, so there's no new writing interface to learn.

## RFP Response

### The draft outline

Start from the **Proposal Documents** tab of a pursuit:

<Steps>
  <Step title="Click New Document, then Automatic Proposal Draft">
    Sweetspot analyzes the solicitation (Sections C, L, and M) to build the outline.
  </Step>

  <Step title="Specify the files to use">
    Pick the pursuit files and Library files the draft should be based on, and add custom instructions if you have them. Then click **Generate Draft**.
  </Step>

  <Step title="Review the outline">
    Sweetspot produces a compliance-mapped outline. Each passage has a content plan: what the passage should be about, what it should look like, and the specific content Sweetspot AI should reference.
  </Step>
</Steps>

<img alt="A proposal draft outline with per-passage content plans" />

Work the outline before drafting: edit a passage's instructions, regenerate it, or restructure sections until the plan matches how you want to respond. This is the highest-leverage review point, since every downstream word follows the plan.

### Write the full draft

Once the outline looks right, click **Write Full Draft For Me** at the top, optionally with instructions for the draft (the more detailed, the better). Sweetspot writes a new document from the outline, pulling from your Library so content uses your actual past performance, with citations back to the solicitation documents.

When the draft is ready, click **Move to Sweetspot Editor** to open it as a proposal document.

<img alt="The Sweetspot proposal editor with the Sweetspot AI panel" />

### The editor

The editor works like Word: a formatting ribbon, comments, tables, search and replace, and version history. The **Sweetspot AI panel** on the right answers questions about the document and makes edits for you. Select text first to point the AI at a specific passage, or just describe the change you want anywhere in the draft.

<img alt="Sweetspot AI converting a bullet list into a table with a reviewable change" />

Highlight any text and tell Sweetspot AI what you want done with it, for example:

* "Make this paragraph more concise."
* "Reference my \[XYZ] past performance more here."
* "Turn this list into a table."
* "Define every acronym on first use in this section."
* "Rewrite this to speak directly to the Section M evaluation criteria."

Every edit comes back as a suggested change you **Accept** or **Reject**, so nothing changes in the document without your approval.

### Review and export

Run AI proposal review to catch compliance issues before your color team does, and use **Review Proposal** in the [Requirements tab](/pursuits/workspace#requirements) to check the draft against every requirement.

To export, go to **File → Export to Word**, or choose **Export with Word Template** if you have a [Word Template](/library/organization-library) in your Library: your branding, headers and footers, and table of contents are applied automatically.

## RFI Response

RFI and sources sought responses are how you get in front of an agency before the RFP exists: they shape the requirements, signal your capability, and often determine who's considered when the solicitation drops. They're shorter than RFPs and come in volume, so speed matters. Sweetspot gives you two routes.

### Route 1: the same flow as an RFP

Use **Proposal Documents** exactly as you would for a full proposal: create an Automatic Proposal Draft from the RFI's files, then review the outline, write the draft, and export.

Choose this route when the RFI is very long or when several people are contributing sections.

### Route 2: draft it in chat with Ask AI

For most RFIs, the faster route is the pursuit's **Ask AI** tab:

<Steps>
  <Step title="Ask for the draft">
    Ask the AI to "Start a draft RFI for this opportunity."
  </Step>

  <Step title="Answer its questions">
    The AI asks questions to get your context: which capabilities to emphasize, relevant past performance, whether you're responding as a prime or sub, and anything else the response needs.
  </Step>

  <Step title="Review the outline, then the draft">
    It writes an RFI outline first, then drafts the full response, all in chat, grounded in the RFI documents and your Library. Ask for changes the same way you'd edit any chat response: "shorten section 2", "add our \[XYZ] past performance."
  </Step>

  <Step title="Turn the response into a deliverable">
    <img alt="Chat response icons including Create Proposal Document and Export to Word" />

    Use the icons under the finished response.
  </Step>
</Steps>

* **Create Proposal Document** turns the response into a proposal document, so you can polish it in the editor like a long RFP, with tracked AI edits and Word export.
* **Export to Word** downloads it directly when the response is ready as is.

<Tip>
  Sources sought responses follow the same pattern: ask Ask AI to draft the capability statement response, and it pulls your certifications, set-asides, and past performance from your Library and company profile.
</Tip>

<Tip>
  Drafts are built from your [Organization Library](/library/organization-library), so keep past proposals, excerpts, and key personnel current. See the [Prompting Guide](/prompting-guide) for pointing the AI at the right folders and files.
</Tip>


# Companies & Contacts
Source: https://docs.sweetspotgov.com/pursuits/companies-and-contacts

Built-in CRM for the companies and people behind every opportunity

Sweetspot includes a built-in CRM for the companies and contacts behind your opportunities: primes, subs, teaming partners, competitors, and the people at each.

## Companies

Find your companies under **Library → Companies**, and link them to the pursuits they're involved in. Each company profile can capture:

* **Legal & identifiers**: legal name, DBA name, aliases, UEI and CAGE codes (which link the profile to federal data), and parent company for ownership links.
* **Relationships**: relationship status (teaming partner, competitor, customer, and so on), priority, primary contact, and last-contacted date, so account management context lives with the record.
* **Compliance**: SAM registration status and expiration date, business size, socioeconomic status, and facility clearance, the facts you check before teaming.
* **Capabilities**: a capabilities summary, core capabilities, primary industry, NAICS and PSC codes, contract vehicles, and set-asides, so you can search your network by what companies actually do.
* **Departments and people**: organize a company into departments and attach the people you know there; Sweetspot also suggests team members it finds.
* **Basics**: website, LinkedIn, employee range, and a description.

An **activity timeline** tracks interactions and mentions, so the history is there when the recompete comes around.

## Contacts

Track the people under **Library → People**. Contacts connect to their companies and to pursuits, so you always know who your points of contact are on any given bid. Each person can capture:

* **Identity**: full name plus preferred name, suffixes, and aliases, so search finds "Bob Smith" even when the resume says "Robert."
* **Contact details**: email and alternate email, mobile and office phone, and LinkedIn.
* **Affiliation**: the company or agency they belong to, their department, and their title. People are either **internal team members** or **external contacts**.
* **Engagement**: relationship status, priority, tags, and last-contacted date for external contacts.
* **Resume**: attach an existing resume or upload one; for internal people, this is what the AI draws on when matching personnel to solicitation requirements.

## Auto-extraction and import

Sweetspot can extract company and contact information from documents: upload a capability statement or an email thread and the AI populates the profile. You can also bulk-import companies and contacts from spreadsheets and merge duplicates.

## Common uses

* **Teaming**: search your partner network by capability when the capability matrix shows a gap.
* **Recompetes**: the incumbent's profile carries your interaction history from the last cycle.
* **Turnover**: when an employee leaves, their relationship history stays in the record.

<Tip>
  A company's award history lives in its [Federal Market Intelligence](/intelligence/federal-market-intelligence) recipient profile; the CRM profile holds your private relationship notes. Both link from the same company.
</Tip>


# Pursuits
Source: https://docs.sweetspotgov.com/pursuits/dashboard

Track every opportunity from discovery to award

<img alt="A pursuits pipeline dashboard with value, win rate, velocity, and stage breakdown" />

Pursuits is your pipeline: every opportunity your team has committed to working, tracked from discovery through award, with views, analytics, and an AI that answers questions about your pipeline.

## Views

Four views of the same pipeline:

* **Dashboard** — pipeline KPIs: value by stage, win rates, deal velocity, and stage conversion.
* **List** — a sortable, filterable table of every pursuit with custom columns.
* **Board** — a kanban view; drag and drop pursuits between stages.
* **Calendar** — every deadline and submission date on a calendar.

<img alt="Board view of pursuits" />

## Stages and fields

Pursuits move through customizable stages from pursuit to proposal to won or lost. You can add custom fields, custom dropdowns, and required-per-stage fields, so a pursuit can't advance without the information your process requires.

## Alerts and activity

The built-in alerts system flags pursuits that are overdue, due this week, or going stale, and an activity feed shows what changed across your pipeline and who changed it.

## Ask AI about your pipeline

Ask questions about your pipeline in plain language, right from the Pursuits Dashboard:

```
Which opportunities close this quarter?
What is our win rate by agency?
```

## Team collaboration

Every pursuit has owners and collaborators, threaded notes, file attachments, and task assignments with deadlines. When a pursuit moves from capture to proposal, its notes, win themes, and stakeholder intel stay attached and available to the drafting team.

<Note>
  Sweetspot can also sync pursuits bi-directionally with Salesforce. See [Integrations](/settings/integrations).
</Note>

## Go deeper

Each pursuit is a full workspace with files, notes, AI chat, solicitation shredding, and proposal documents. See [The Pursuit Workspace](/pursuits/workspace).


# The Pursuit Workspace
Source: https://docs.sweetspotgov.com/pursuits/workspace

Everything about one opportunity in one place: files, requirements, drafts, and AI

Every pursuit has its own workspace. Each tab handles one part of the capture-to-proposal flow: understand the solicitation, decide and plan, prove compliance, and draft the response.

## Overview

The Overview is where you check where a pursuit stands. It shows:

* The **stage pipeline** (for example Assessment → Pursuit → Pre-Proposal → Proposal → Submittal → Won/Lost), estimated value, and PWin
* Opportunity details pulled from the solicitation: solicitation number, agency, office, NAICS, and notice type
* Notes and files, plus tabs for activity, discussion, and win themes
* Quick actions: edit, manage access, link to Salesforce or GovWin, and record a **No Bid**

Use it as the starting point for status checks and stand-ups: one look tells you the stage, the deadline, and what's changed recently.

<img alt="The pursuit Overview tab with stage pipeline and opportunity details" />

## Wiki

The Wiki is an AI-generated snapshot of everything in the solicitation, so anyone on the team can get up to speed without reading 300 pages of RFP. Sweetspot builds it from the pursuit's files and organizes it into sections like scope of work, technical requirements, key constraints and risks, evaluation factors, and proposal instructions.

* Every page lists the **files used**, so you can verify claims against the source
* **Regenerate** rebuilds the wiki when amendments or new documents arrive

Use the Wiki at the start of a pursuit for bid/no-bid context, and when new teammates join mid-pursuit. For your own notes and strategy, use [Notes](#notes).

<img alt="The pursuit Wiki tab with an AI-generated summary of the solicitation" />

## Ask AI

A chat scoped to this pursuit. The AI sees the pursuit's files, requirements, and notes, and answers with citations. Suggested Playbooks and the workspace rail (pursuit files, Library files, citations) sit alongside the conversation. See [Ask AI](/ai/ask-ai).

Use it for questions the Wiki doesn't answer directly: "What are the Section L page limits?", "Does the SOW require a SCIF?", "Compare amendment 2 against the original deadline."

<img alt="The pursuit Ask AI tab with suggested Playbooks and workspace files" />

<Note>
  Pursuit chats are different from the **Ask AI** tab in the left sidebar. A chat started here has all the context from the pursuit's files, and it lives in this tab: it won't appear in the broader Ask AI chat list. If you're looking for a conversation about a pursuit, open that pursuit's Ask AI tab.
</Note>

## Tasks

Task tracking for the pursuit, so ownership and deadlines are explicit. Each task has a status, priority, assignee, and due date, plus comments with @mentions to notify teammates. The calendar view lays every deadline out by date.

Use Tasks to run the proposal schedule: color team dates, volume owners, question submission deadlines, and the small to-dos that fall through the cracks in email.

<img alt="A task in the pursuit Tasks tab with status, priority, assignee, and comments" />

## Concierge

Route pursuit work to the Sweetspot team as a managed service. Available on select plans.

## Files

The pursuit's document repository: solicitation documents, amendments, and working files.

* Organize with **folders** and **tags**, and search within a folder
* Upload files directly, or let integrations (like GSA eBuy) attach them automatically
* Copy files worth reusing into your [Library](/library/organization-library) so future pursuits benefit

Everything here is what the AI reads: the Wiki, Shred, and pursuit chats all work from these files, so keeping Files current keeps everything else current.

<img alt="The pursuit Files tab with documents, folders, and tags" />

## Notes

Rich-text notes with formatting, sharing, and export to Word. AI-drafted content, like win themes, lands here with citations back to the source files.

Use Notes for content that's pursuit-specific but not part of the proposal document itself: win theme drafts, call notes, and internal strategy.

<img alt="The pursuit Notes tab with a draft win themes note" />

## Shred

Shred is the foundation the matrices are built on. Sweetspot parses each solicitation document into categorized citations:

* **Administrative** requirements
* **Evaluation Factors** (Section M)
* **Proposal Instructions** (Section L)
* **Task Areas** (Section C / SOW)
* **Shall / Will / Must** statements

Shred runs automatically when a pursuit is created and solicitation files are added, so by the time you open this tab the documents are already parsed.

Each citation is highlighted in the source document, side by side with the citation list, so you can verify every extracted requirement in context before it flows downstream. Filter citations by category, and archive documents you've finished analyzing.

Review the Shred output early: it's what populates Requirements and Capabilities.

<img alt="The Shred tab with a parsed solicitation and categorized citations" />

## Capabilities

The capability matrix answers one question: **can we actually do this work?** Every capability extracted from the shred becomes a row, scored against your organization, so you can make the bid/no-bid call on evidence and spot the gaps that need teaming partners.

<img alt="The Capabilities tab scoring capability fit, risk, and past performance" />

Click **Run AI Fill** to score every cell from your Library and company profile. To redo a subset, select cells and use **Run AI Fill on selected cells**. Open a row's details to see the reasoning and sources behind a score.

**Build the columns your process needs.** The matrix starts without columns; you decide what to score. Open the **Columns** panel and either:

* Add a **suggested column** with the **+** next to its name. These cover the dimensions teams typically need, like Capability Fit, Risk Level, and Past Performance Relevance.
* Create a **custom column**: describe the data you want to extract from the solicitation shred. Name the column, give the AI instructions, and pick the answer structure: free-form text, select one, or select many, with your own answer options or options pulled from your Library. For example, add a custom column called "Past Performance Relevance Mapped" and it grades each capability (like "Strong fit") and links the Library file it based the grade on, so you can check the source yourself.

**Act on the gaps.** Rows where fit is weak or risk is high are your teaming checklist: search [Companies](/pursuits/companies-and-contacts) and your partner library for coverage. Export the matrix when you need it in a deck or a teaming discussion.

## Requirements

The compliance matrix answers the other question: **are we meeting every requirement the government set?** Every requirement from the shred becomes a compliance item with full traceability from RFP text to your proposal response.

<img alt="The Requirements tab with compliance statuses per requirement" />

Each item carries:

* The raw requirement text and its **RFP section and page**
* A **category** (Administrative, Evaluation Factors, Proposal Instructions, Task Areas, Deliverables)
* An **assignee** and **due date**
* A **status**: Not Reviewed, Not Compliant, Partially Compliant, or Compliant, with reasoning
* The **proposal section, page, and excerpt** where your response addresses it

**Work the matrix.** Assign an owner and due date to every requirement early; the sidebar filters (by assignee, status, category, and file) then become each person's work queue. Statuses can be set manually as sections get written.

**Check the draft automatically.** Click **Review Proposal** and the AI reads your proposal documents against every requirement: it sets the compliance status, explains its reasoning, and links the proposal excerpt where the requirement is addressed. Run it before every color team to catch gaps while there's still time to fix them.

**Export to Excel.** Use **Export to Excel** for color team reviews, compliance checklists for teammates outside Sweetspot, or submission records. Bulk actions handle status and assignee changes across many items at once.

## Proposal Documents

The proposal editor workspace: outlines, drafts, forms, and spreadsheets. **New Document** gives you four starting points:

* **Automatic Proposal Draft**: Sweetspot analyzes the solicitation (Sections C/L/M) and builds a compliance-mapped outline with section-by-section content plans. The fastest path to a compliant first draft.
* **Manual Proposal Outline**: define the outline structure yourself with the drag-and-drop builder; the AI generates content plans for each section. Best for reusing proven structures.
* **Template**: start from a Library template in the Word-like editor; Sweetspot fills placeholders, tables, and boilerplate as you draft. Good for RFIs, capability statements, and letters.
* **Blank Document**: start from a clean page and write at your own pace, with the AI generating and rewriting text as you ask for it.

<img alt="Creating a new proposal document with four starting points" />

See [Proposal Engine](/proposals/proposal-engine) for the full drafting pipeline.

## Typical sequence

<Steps>
  <Step title="Qualify">
    Read the Wiki for the full snapshot of the solicitation, then take your open questions to Ask AI: who the incumbent is, whether you're eligible for the set-aside, what clearances or facilities are required, how the evaluation factors are weighted, and where your gaps are. Record the bid decision and strategy in Notes.
  </Step>

  <Step title="Verify the Shred">
    Shred runs automatically on the solicitation files. Open the tab and spot-check the extracted citations against the highlighted source text, especially evaluation factors and proposal instructions, since Requirements and Capabilities are built from them.
  </Step>

  <Step title="Assess your fit">
    In Capabilities, add the columns you score bids on and run AI Fill. Review the weak-fit and high-risk rows: those are the capabilities to shore up with teaming partners before you commit resources.
  </Step>

  <Step title="Draft the proposal">
    In Proposal Documents, start from an Automatic Proposal Draft (or a manual outline or template), then work section by section with the AI pulling from your Library.
  </Step>

  <Step title="Check compliance">
    In Requirements, give every requirement an owner and a due date so the matrix becomes each person's work queue. As sections are written, mark compliance manually, or run Review Proposal to check every compliance item against the draft automatically. Re-run it before each color team and export to Excel for the review.
  </Step>
</Steps>


# Quickstart
Source: https://docs.sweetspotgov.com/quickstart

Set up your account, run your first search, and qualify an opportunity

This walkthrough covers the core Sweetspot workflow: give the AI context about your business, find an opportunity, qualify it, and start pursuing it.

<Steps>
  <Step title="Add your business info in Settings">
    Go to **Settings** and fill in your organization's basics: name, website, description, UEI and CAGE code, and address under **Overview**, and your NAICS codes, PSC codes, and set-asides under **Classifications**. These drive opportunity matching and AI recommendations. See [Settings](/settings/settings).
  </Step>

  <Step title="Build your Library">
    Go to **Library → Files** and upload the documents that describe your business: past proposals, capability statements, key personnel resumes, and pricing sheets. Sweetspot indexes them and uses them in search results, capture briefs, and drafts.

    <Tip>
      If your content lives in SharePoint, connect it under **Settings → Integrations** and sync folders directly instead of uploading files one at a time. See [Integrations](/settings/integrations).
    </Tip>
  </Step>

  <Step title="Search for opportunities">
    Open **Search** and describe what you're looking for in plain language, for example:

    ```
    IT modernization contracts with the Navy over $5M
    ```

    Sweetspot searches federal contracts, federal grants, DIBBS, SBIR/STTRs, and state and local sources. Use filters for structured criteria like NAICS codes, set-asides, or solicitation numbers. See the [Search guide](/discover/search).
  </Step>

  <Step title="Qualify with Ask AI">
    Open any opportunity and click **Ask AI**. Ask about requirements, evaluation criteria, incumbents, or fit:

    ```
    Does my company qualify for this opportunity? What gaps do we have?
    ```

    Answers are cited to the solicitation documents. Generate a capture brief for a structured bid/no-bid recommendation.
  </Step>

  <Step title="Add it to your Pursuits">
    To pursue an opportunity, add it as a **Pursuit**. It enters your pipeline, where you can assign owners, track tasks and deadlines, take notes, and move it through stages from pursuit to proposal to won or lost. See [Pursuits](/pursuits/dashboard).

    Not ready to commit? [Bookmark it](/discover/bookmarks) instead.
  </Step>

  <Step title="Automate your discovery">
    Save your best queries as [Saved Searches](/discover/saved-searches) to get alerted when new matches post, and set up [Monitors](/intelligence/monitors) to watch specific agencies, competitors, and opportunities. New activity lands in your [Inbox](/intelligence/inbox).
  </Step>

  <Step title="Draft the proposal">
    Sweetspot automatically shreds the solicitation into requirements when files are added to a pursuit. Review the compliance matrix under **Requirements**, then generate an outline and first draft in **Proposal Documents**. Export to a branded Word document when you're ready. See the [Proposal Engine guide](/proposals/proposal-engine).
  </Step>
</Steps>

## Next steps

<CardGroup>
  <Card title="Prompting Guide" icon="lightbulb" href="/prompting-guide">
    Get better results from search, chat, and the proposal writer.
  </Card>

  <Card title="Federal Market Intelligence" icon="chart-line" href="/intelligence/federal-market-intelligence">
    Research incumbents and recompetes before you bid.
  </Card>

  <Card title="Invite your team" icon="users" href="/settings/settings">
    Add teammates and set roles under Settings.
  </Card>

  <Card title="Connect your tools" icon="plug" href="/settings/integrations">
    Microsoft 365, Salesforce, GovWin, GSA eBuy, and Slack.
  </Card>
</CardGroup>

Stuck on anything? Click **Live Support** in the console to chat with the Sweetspot team directly.


# Integrations
Source: https://docs.sweetspotgov.com/settings/integrations

Connect Sweetspot to Microsoft 365, Salesforce, GovWin, GSA eBuy, and Slack

Organization admins manage three integrations under **Settings → Integrations**: Microsoft 365, Salesforce, and GovWin. GSA eBuy and Slack are set up from the surfaces that use them, covered at the end of this page.

## Microsoft 365

<img alt="Microsoft 365 integration" />

Connect your Microsoft account and Sweetspot AI can work across your Microsoft stack:

* **Outlook**: draft, review, and send email; manage meetings.
* **Word**: create documents directly.
* **PowerPoint**: generate presentation decks.
* **OneDrive & SharePoint**: browse and use your files in chat and the Library.
* **Teams**: send messages.

Sweetspot supports **Commercial, GCC, and GCC High** cloud environments. Admins can grant org-wide consent, choose Sweetspot-managed or custom (Entra) credentials, and control which **capabilities** are available in chat for the integration.

### SharePoint

SharePoint access comes with the Microsoft 365 integration; there's no separate connection to set up. Once connected, browse your SharePoint files from **Library → SharePoint** and use them in chat and drafting. Admins can restrict access to selected sites.

## Salesforce

Bi-directional sync between Sweetspot pursuits and Salesforce. Stage mapping keeps both systems current, so your team can work in Sweetspot while reporting stays in the CRM. Admins configure the Salesforce login domain and complete the sign-in to connect.

## GovWin

Connect your Deltek GovWin account to search and browse GovWin opportunities inside Sweetspot chat, analyze them with Sweetspot AI (incumbents, awarded contracts, buyer contacts), work through your GovWin bookmarks, and import opportunities into Sweetspot as pursuits with their solicitation documents.

## GSA eBuy

Set up from **Inbox**: forward your eBuy RFQ notification emails to your organization's unique Sweetspot ingest address, and every RFQ lands in your [Inbox](/intelligence/inbox) automatically with its attachments.

<Note>
  GSA eBuy ingestion is available on request.
</Note>

## Slack

Connect your Slack workspace from the Inbox, then pick a channel per monitor in its Subscribers step to post alerts there. Your Slack identity can be managed (or disconnected) under your user settings' **Account** tab. See [Monitors](/intelligence/monitors).


# Settings
Source: https://docs.sweetspotgov.com/settings/settings

Account, organization, team, and notification management

Sweetspot has two settings surfaces: **organization settings**, opened from **Settings** in the sidebar, and **user settings**, opened by clicking your user profile.

## Organization Settings

### Overview

Your organization's identity: the basic facts Sweetspot and your teammates reference everywhere.

* **Name**, **website**, and **description** (the description tells the AI what your company does)
* **UEI** and **CAGE code**
* **Point of contact**: name and email
* **Address**: street, city, state, and ZIP

### Classifications

The codes that define your market. Keep these current: they drive opportunity matching and AI recommendations.

* **NAICS codes**
* **PSC codes**
* **Set-asides** (8(a), HUBZone, WOSB, SDVOSB, and so on)

### Members

Who's in your organization. Add and remove members and manage seats; role-based access control scopes who can see and edit what, and enterprise organizations can add SSO.

### Integrations

Connections to Microsoft 365, Salesforce, GovWin, GSA eBuy, and Slack. See [Integrations](/settings/integrations).

### Pursuit Preferences

How Sweetspot extracts and organizes requirements when it shreds a solicitation:

* **Keyword dictionaries** for each requirement category: Administrative (compliance and administrative requirements), Evaluation Factors (criteria for proposal evaluation), Proposal Instructions (instructions for proposal preparation), and Task Areas (statement of work and deliverables). Tune these so shredding flags the language your team cares about.
* **Pursuit colors** for the categories, so the same color coding follows requirements across the workspace.

### CUI Settings

Organizations handling Controlled Unclassified Information can turn on CUI compliance mode, which disables features that cross the CUI boundary. If your organization has compliance requirements, talk to your Sweetspot representative about the right deployment.

### Search Settings

Your organization's **default opportunity type**: which search page (Federal Contracts, Federal Grants, DIBBS, SBIR/STTRs, or State & Local) the Search icon opens first.

## User Settings

Click your **user profile** to open your personal settings, organized into three tabs.

### Account

Your name, email, and avatar.

### Admin Dashboard

Member management for organization admins: invite teammates, revoke pending invites, and configure enterprise SSO. Non-admins see this tab as **Organizations**, showing their memberships.

### Preferences

Your notification preferences, personal to you rather than organization-wide. Toggle notifications for things like Saved Search results, monitor alerts, pursuits and tasks assigned to you or due soon, comments on your tasks, replies to your comments, and chats shared with you.

## Changelog

New features are announced at [console.sweetspotgov.com/changelog](https://console.sweetspotgov.com/changelog) and via in-app notifications.