Everything below lives in the rail's Workspace group.

Work the compliance matrix

The shred turned every requirement it found in the solicitation into a row. Open Compliance matrix from the rail and each row has a status you can set to Open, Addressed, or N/A, plus an Addressed by section picker mapping it to the section that answers it. The rail shows a running count and a mini-bar, so you can see coverage without opening the drawer.

Two ways to close a row. Write the answer yourself in the mapped section, then mark the row Addressed. Or click the AI button on the row, which drafts the mapped section to address that requirement; the same thing is available from inside the section editor as Address requirement…, which rewrites a highlighted selection to answer a specific row. Either way, the matrix stays fully editable, and it ships in the export as a compliance checklist so an evaluator can trace every requirement to a page.

Our advice: work the matrix before the review rounds, not after. Reviewers should be reading a proposal that's already compliant, so their attention goes to persuasion.

Review rounds and color teams

Color teams are the traditional proposal review sequence, and Cairn uses the standard names: Blue for strategy and win themes, Pink for the first full draft, Red for compliance and competitiveness, Gold for the final executive review, Green for pricing and cost, and White for lessons learned.

Open Review rounds, pick a color team, give the round a due date, and click Add round. A round moves through Planned, In progress, and Complete. Reviewers work in the section editor's Review mode: highlight a span of text and pin a comment to it. Each comment belongs to the active round and carries a severity: Must (blocking), Should, or Consider. Comments can be resolved as they're handled, and the outline rail shows an open-comment count on each section so nothing hides.

We'd suggest keeping the severities honest. A Must is something that would lose the bid or fail compliance. If everything is a Must, the submission package will block on all of it, which is the right behaviour but a miserable afternoon.

AI review

AI review runs the same exercise with the AI as reviewer. Choose a pass: Pink checks a first draft for compliance and completeness, Red reads as a skeptical evaluator for persuasiveness and win themes, and Gold does a final executive polish. Click Run review and the AI reads each written section and leaves anchored comments in a new round of that color, pinned to the exact text like a human reviewer's. If it finds nothing, it says so rather than inventing feedback.

One person gets the benefit of a review team. That's the point of this feature, and it's why we run it on our own proposals in the order Pink, Red, Gold, with a real human reading between passes.

Turning comments back into text

Two tools convert feedback into edits, and both leave you in control.

AI: propose a fix works on a single comment. It drafts a replacement for the commented span, and accepting it writes a new version of the section.

Evaluate comments & regenerate works at section scale. The AI reads the section's current text plus every open comment and rewrites the whole section. The result appears as a proposed revision beside the current text, and nothing changes until you click Accept revision (written as a new AI version) or Reject. If you edit the section after generating, Cairn asks you to re-run, so the AI is always working from your latest text rather than a stale copy. The button is disabled while the editor has unsaved changes for the same reason.

The SF-424 cover sheet

For grant submissions, the rail has an SF-424 cover sheet entry. The SF-424 is the standard cover form for federal grant applications. Cairn auto-seeds it from your company profile and the opportunity, and it renders as the cover sheet in your export and in any Grants.gov filing. Re-populate from profile refreshes the auto-filled fields after you've changed the profile.

Two fields deserve a note. The Assistance listing (CFDA) number is the five-digit code identifying the exact federal grant program, such as 93.243; it's on the funding opportunity, and the CFDA title is that program's official name. And estimated funding has three parts: federal is what you're requesting from the agency, applicant is your organization's own contribution or cost-share match, and total is the two combined. Enter your match even when the program doesn't require one.

The submission package

Submission package assembles everything into one ready-to-file bundle, with a checklist standing between you and a mistake.

The ZIP contains the proposal as both DOCX and PDF, the compliance checklist, the SF-424 cover sheet for grants, the bid-decision summary from the scorecard, your approved past performance, your attachments under an attachments folder, and a manifest documenting completeness. Download package gets you the same ZIP the header's Export button does.

Attachments is where budgets, narratives, resumes, and letters go, each in a slot (Narrative, Budget, Resume, SF-424 form, or Other). The feature we're proudest of here is small: you can declare a required item before its file exists, with Add required item. Readiness will then block until something is actually attached, which is what stops a known-missing document from being forgotten under deadline pressure.

The readiness checklist

Two severities, and only one of them stops you.

Blocking checks, shown in red, must be resolved before final approval: open compliance items, empty sections, missing required SF-424 fields, required attachments with no file, unresolved Must comments, and a missing solicitation. Warnings, shown in amber, are advisory: Should and Consider comments, pending AI revisions you haven't accepted or rejected, and no P-win recorded on the scorecard.

The header's readiness chip summarises the same list, so you can watch blockers count down without opening the drawer.

Final approval

Approve all documents & info for submission is an explicit human sign-off. You're attesting that every document and field has been reviewed. The button stays disabled until every blocking check passes, and until then it explains what's still open. Approving freezes the submission at Final and notifies the pursuit's assignee that it's ready to submit. A submission moves through four states: Draft while it's being written, Final once approved, Submitted once filed, and Abandoned if the pursuit is dropped.

Filing with Grants.gov

For grants on the Enterprise plan, Cairn can file directly through the Grants.gov System-to-System (S2S) API, transmitting the SF-424 generated from your cover sheet plus every attached file. After approval, a Submit to Grants.gov button appears; it shows a confirmation listing the passed checks and the attachments about to be transmitted, and Confirm & submit sends it. The tracking number and processing status (Received, then Validated) then appear in the package, with a refresh control.

This needs setup, because Grants.gov requires each applicant to submit through its own Authorized Organization Representative. The credential lives on the company profile, in the Grants.gov submission card:

1. In Grants.gov, have your AOR request S2S access and obtain your organization's S2S certificate.
2. Export it as a PKCS#12 file (.pfx or .p12) with a password. This is the mTLS client certificate tied to your AOR-enrolled account.
3. Upload the file and enter its password in Cairn. It's stored encrypted and never shown again; upload a new file only to replace or renew it.
4. Select the Test environment first, validate a real submission end to end, then switch to Production.

The Test environment is Grants.gov's applicant acceptance environment: it validates a real application without filing it. Only switch to Production once a Test submission has succeeded, because Production files for real. Enable S2S submission is the master switch for your company; automated filing is attempted only when it's on and a certificate is uploaded, and turning it off pauses automated filing without deleting your certificate. The sender ID field is optional and only needed if Grants.gov or the agency gave you one.

Until S2S is set up, or on plans without it, Cairn gives you guided manual instructions instead: download the package, sign in to Grants.gov Workspace as the AOR, complete the SF-424 from the cover sheet in the ZIP, upload the narrative and attachments, then record the tracking number in the pursuit's notes. Contract submissions follow the solicitation's own instructions: upload the DOCX or PDF and the attachments where the agency says, record the confirmation in the pursuit's running notes, and move the card to Submitted.

That's the whole path, from a search result in the pipeline post to a filed application. The pieces that feel strict, the approval gate and the blocking checks, are the ones that have saved us the most.

Questions people ask

What blocks final approval?

Six things: open compliance items, empty sections, missing required SF-424 fields, required attachments with no file, unresolved Must comments, and a missing solicitation. Should and Consider comments, pending AI revisions, and a missing P-win only produce amber warnings. The approve button becomes available the moment every blocking check passes.

Can Cairn submit a contract proposal to SAM.gov for me?

No. Automated filing is available only for grants, through the Grants.gov System-to-System API on the Enterprise plan. For a contract, Cairn assembles the package and gives you step-by-step instructions to submit it the way the solicitation specifies, then you record the confirmation and move the pursuit to Submitted.

What is in the exported ZIP?

The proposal as both DOCX and PDF, the compliance checklist, the SF-424 cover sheet for grants, the bid-decision summary from the scorecard, your approved past performance, your attachments in a folder, and a manifest documenting completeness. Download package in the submission drawer and Export in the header produce the same bundle.

Do I need an AOR certificate to use the submission package?

No. The package, its readiness checklist, and final approval work on every plan that includes the workspace, and manual filing instructions are always available. The AOR-enrolled S2S certificate is needed only for automated Grants.gov filing, which is an Enterprise feature, and it is uploaded on the company profile under Grants.gov submission.

The workspace is part of the Professional and Enterprise plans. This post is the front half of the workflow; part two covers compliance, review rounds, and the submission package.

Where a proposal comes from

Most proposals in Cairn start themselves. When a pursuit advances into the Proposal in Progress stage, Cairn queues an automatic first draft: it creates the submission, picks an outline for a contract or a grant, runs a compliance pre-pass over the solicitation's attachments, and drafts every section from the gathered pursuit data and your company profile. The workspace header shows AI draft queued and then AI drafting sections while that runs, and each section fills in as it finishes. If a draft fails partway, you get a Retry AI draft button rather than a silent gap. Once it's done, Regenerate draft will redo every section into a new AI version; your current content stays in version history.

The manual route works too. Open a pursuit from the pipeline, go to its Tools tab, and under Proposal click Create proposal. Either way you land in the workspace at Deliver › Proposals, where the list shows every submission with its status chip, its due date, and a progress bar reading "6 of 9 sections drafted".

The workspace at a glance

A sticky header carries the title, the status chip (Draft, Final, Submitted, or Abandoned), the due date with a days-left count, a readiness chip that reads Ready, Not ready, or the number of blockers, and an Export button that downloads the whole thing as a ZIP.

Down the left is the rail, in three groups. Outline lists every section with a state dot (Drafted, AI drafting now, Outline seed, or Not started) and a word count. Workspace opens the compliance matrix, AI review, review rounds, the SF-424 cover sheet for grants, and the submission package. Sources opens the solicitation and shred card and your context documents.

The middle is the document itself, one continuous canvas with a Draft with AI strip at the top and each section as a numbered block.

Set up sources first

The first time you open a fresh proposal, the Sources drawer opens on its own with the title Set up sources. Take the thirty seconds. This is where you tell the AI what it may read.

Context documents are reference material from your company: past proposals, writing samples, standard language. Each file is uploaded once for your whole company (it also appears in the Library under Content as a "Context document"), and each new proposal just ticks the boxes for the documents it should draw on. Then you give every selected document one of two roles:

• Content source: the AI draws facts, capabilities, and evidence from it.
• Style guide: the AI matches its tone, structure, and formatting, and won't copy its facts.

That split is the part we'd ask you to think about. A past proposal that won is a great style guide and a risky content source, because its facts belong to a different customer and a different year. A current capability statement is the reverse. Marking the roles keeps the AI from quoting last year's staffing plan into this year's bid.

The whole set of context documents does something broader: it teaches Cairn your company's voice profile, a distilled picture of your tone, structure, and phrasing. The profile refreshes when a document finishes processing, and there's a Rebuild voice profile button in the card. It captures style only. The AI never copies facts or client names out of your samples.

If you have nothing to add yet, Continue without documents closes the drawer and remembers that you chose to. You can come back any time from Context documents in the rail. One detail: a context document that isn't approved in the Library shows a Paused chip here, and the AI won't read it until someone approves it.

Shred the solicitation

Now the government's document. In the same drawer, the card titled Solicitation analysis ("shred") takes the RFP or funding notice as a PDF, DOCX, TXT, or MD file. Upload it with Upload the RFP / funding notice, then click Build matrix & outline (shred).

"Shred" is capture jargon for tearing the document apart line by line so that no requirement is missed. The run produces two things: a compliance matrix of every requirement it found, and a proposed section outline for your response. Don't upload your own reference material here; that belongs in context documents, and mixing the two is the most common way to get a confused outline.

The shred deliberately pauses partway through and shows you the proposed outline. Approve it as-is or edit the section list first. Nothing is written to your proposal until you approve, and if a teammate is assigned to the pursuit, they're notified that an outline is waiting for review.

No solicitation to shred? Start from a standard outline instead. The picker offers Federal contract (UCF volumes), which seeds Executive Summary, Technical Approach, Management Approach, Staffing & Key Personnel, Past Performance, and Cost/Price Volume, or Federal grant (narrative + SF-424), which seeds the standard grant narrative sections from Statement of Need through Budget Narrative. Each section arrives with a short purpose note as its first version (the rail marks it Outline seed), and you can add, rename, reorder, or remove sections from the block menu.

Drafting

There are three ways to get words into a section, and they compose.

Draft with AI on a single section writes it grounded in your company profile, your approved past performance, your saved content, and the context documents you selected. The result is saved as a new version tagged AI, so machine text stays distinguishable from your edits and you can roll back to any earlier version from History.

Draft N sections, in the strip at the top of the canvas, first-drafts every section that's still empty. Sections you've already written, or that have a pending AI revision waiting for your decision, are skipped. It will not overwrite your work. That rule is absolute, and it's why the button says how many sections it's about to touch.

Edit & comment opens the section editor. In Edit mode you write Markdown in a plain textarea, and the toolbar above it holds the AI actions for a selection: highlight some text and choose Improve, Tighten, Expand, or Align to win themes, or pick a requirement from Address requirement… to rewrite the selection so it answers that row of the compliance matrix. The AI rewrites just the selection and drops it back in place, and nothing is saved until you click Save.

Each drafted section shows a footer with its version number, whether it's an AI draft or human text, its word count, how many requirements it covers, and when it was last updated. Restoring an older version is one click from History.

What the AI reads, and what feeds it

The quality of a draft tracks the quality of what the AI can see. Three things feed it. Your company profile, especially the capability narrative. Your Library, which holds past performance write-ups (marked prime or subcontractor, because evaluators weigh the two differently) and approved reusable content: win themes, boilerplate, resumes and bios, strong passages lifted from past proposals, and the context documents above. And the pursuit itself, including whatever the Intelligence tab gathered from the notice.

The Library has one gate worth understanding: a new asset is a draft, and the AI ignores it. Approving an asset with Approve for AI indexes it into your company's private search store, and only then can a draft retrieve and cite it. The badge reads Approved · indexed when that's true. This gate exists so unreviewed language can't find its way into a proposal by accident. A small library of your best language produces better drafts than a large one of mediocre content, so approve selectively.

That gets you from an empty submission to a drafted document with a compliance matrix behind it. Part two picks up from there: working the matrix, running review rounds, and getting the package ready to file.

Questions people ask

Will the AI overwrite something I wrote?

No. Draft with AI saves its output as a new version tagged AI, so your text stays in history and you can restore it. Draft all sections only touches sections that are still empty and skips anything you've written or anything with a pending AI revision. Selection edits change nothing until you click Save.

What's the difference between a context document and the solicitation?

The solicitation is the government's document, the RFP or funding notice you're responding to, and the shred turns it into your compliance matrix and outline. Context documents are your own material, such as past proposals and writing samples, which ground the AI's drafts and teach it your voice. Upload each in its own card; mixing them produces a confused outline.

Which plan includes the proposal workspace?

The workspace, the solicitation shred, AI drafting, and AI review are part of the Professional and Enterprise plans. On Starter the Proposals page shows the feature locked with an upgrade prompt, and after an upgrade it opens right away with your existing pipeline untouched.

Can I start a proposal without an RFP?

Yes. When there is no solicitation to shred, apply a standard outline instead: Federal contract seeds the Uniform Contract Format volumes, and Federal grant seeds the standard narrative sections plus the SF-424 cover sheet. Each section arrives with a purpose note, and you can add, rename, reorder, or remove sections before drafting.

Find it

Open Capture › Find Opportunities. One search covers both federal registries, contract opportunities from SAM.gov and grant funding opportunities from Grants.gov, and the two source pills let you switch either one off. Results are merged and de-duplicated, so a notice that appears in more than one place shows once, keeping its highest fit score, and each result carries a badge for where it came from.

Type keywords for the work you do and search. The Use company profile pill is on by default; it's what ranks the results against your profile rather than returning them in the order the registries did.

The filter bar narrows by NAICS (up to five codes), set-aside (Total Small Business, 8(a), WOSB, SDVOSB, HUBZone), state of performance, and due window (30, 60, or 90 days). Here is the single most important thing to know on this page: those structured filters only narrow SAM.gov results. Grants.gov opportunities don't carry those fields, so grants are matched on your keywords alone, and the filters never constrain them. The page says so in small print, and we'll say it again here because it surprises everyone once.

By default Cairn hides anything whose deadline has passed, plus anything posted more than two years ago. A Grants.gov opportunity can stay "posted" for years, and without the age cut-off stale listings crowd out live ones. Toggle Include expired & very old to see them; those results are chipped Expired or Very old.

Reading a result

Every result has a fit score from 0 to 100 and a Why it fits list explaining it. The score reads your NAICS and PSC codes, your set-aside certifications, your target agencies, and your capability narrative. It's triage, not a verdict.

Results you can't bid are greyed but kept visible, and a high-fit result your certifications bar you from gets an explicit Ineligible flag. We flag rather than hide on purpose. A strong fit score shouldn't nudge you toward work you're not permitted to win, and seeing what you're locked out of is useful information about which certifications would open doors.

The value figure means different things by source. For a contract it's the government's estimated total value of the work. For a grant it's the funding ceiling, the most a single award can be. Either way it's the official estimate, not a promise.

The results line reads something like "12 matches · from ~40 notices". Matches is what you're seeing after de-duplication and your filters. The larger number is the approximate total the sources reported before local filtering.

When something looks worth a closer read, click Add to Pipeline. It lands in Identified, the first stage, ready to qualify.

Two other ways in

Search isn't the only intake. The dashboard's Top Matches card lists recently captured opportunities ranked against your profile, each with an Add button. And on the Professional plan, Capture › Recompete Radar runs a weekly sweep of expiring federal awards in your NAICS codes and target agencies, so you can position for the follow-on before the solicitation drops; Add to pipeline on a candidate creates the pursuit and saves a snapshot of the award with it. There's no free-text "add an opportunity by hand" path. Every pursuit starts from a real notice or a real award, which is what lets the rest of the product read the source document.

The board

Capture › Pipeline is a board: columns are capture stages, cards are pursuits, and moving a card is a decision you're recording. The stages run left to right: Identified, Qualifying, Pursuing, Proposal in Progress, Submitted. After that a pursuit reaches a terminal outcome, Won, Lost, or No-bid, and drops off the active board into a collapsed Closed strip at the end, where Won outcomes get a gold chip. Closed pursuits no longer count against your active-pursuit limit. To reopen one, move it back to an active stage.

A card shows the title, agency, due date, value, a Contract or Grant chip, a fit badge, and an Advance button. Two colour cues are worth learning: a red edge means the deadline is overdue or inside the 14-day critical window, and a gold edge marks a high-fit pursuit. A New chip marks anything added in the last seven days. The Board and Table toggle at the top switches to a sortable table when you'd rather scan by due date or value.

Open the card

Click a card and the pursuit drawer opens with four tabs.

Overview is the opportunity fact grid: agency, type, NAICS, set-aside, response due, a link to the official listing, and the fit score with its Why it fits reasons.

Intelligence is where the AI does some homework. Click Sync and Cairn gathers a summary, key facts, the source documents attached to the notice, points of contact, and links, then shows when it last synced and how many sources it read. This is the tab we'd open first on anything that scored well.

Tools changes with the stage. Early on it holds the bid/no-bid scorecard and the gate review. Once a pursuit reaches the proposal stage it adds the proposal launcher, and at the end the submission package.

Activity is the pursuit's log: every stage move, every note, tagged by who made it (you from the app, the assistant from chat, or the system).

Pinned to the bottom of the drawer is the stage dock: a numbered stepper across the active stages with the current one marked, the terminal chips under Close out, a note box for an optional reason, and an Advance to … button. You can also move a card backward to any earlier stage, and we'd encourage it. A pursuit that needs more qualification than it got, or a solicitation that changed enough to reset your thinking, belongs back in Qualifying. Every move is recorded in the activity log with whatever note you typed, so the board's current state always has a history behind it.

Score the decision

The bid/no-bid scorecard is the structured version of the go/no-go conversation. Six weighted criteria: capability fit, incumbent position, agency relationship, past-performance relevance, resources and capacity, and competitive landscape. You score each from 0 to 10 and write a short justification.

It produces two things worth having. The first is a P-win, a probability of win that's a deterministic weighted roll-up of your scores. The same inputs always produce the same number; there's no black-box model in the middle. Only the criteria you actually score count toward it, and unscored criteria are excluded rather than silently treated as zero. The second is an auditable record of why you chose to bid or pass, which is exactly what you want when someone asks about it a quarter later.

On Professional, Research evidence with AI reads the pursuit and your profile, suggests a score for each criterion with its reasoning, and marks those entries with an AI draft badge. The suggestion is a draft. You own the number, and we built the badge so nobody mistakes the two.

Gate reviews

A gate review is a capture-discipline checkpoint. When a stage has its gate turned on, you must record a bid, no-bid, or defer decision, with an optional rationale, before any pursuit can advance into it. The point is to force a deliberate go/no-go. Without a gate, pursuits drift forward one small step at a time and you discover at proposal time that nobody ever actually decided to bid.

Gates are off by default. Turn them on per stage under Company › Team, on the Pipeline tab. Most teams gate the entry to Pursuing, to Proposal in Progress, or both. Then, in the drawer's Tools tab, the Gate review panel takes the decision and Record gate review files it.

A working rhythm

Three habits keep the board honest. Qualify early and no-bid honestly: a no-bid recorded with a justification is a better outcome than a pursuit that quietly stalls in Qualifying for a quarter. Score the scorecard before the gate review, not during it. And let the terminal stages do their job, because closing pursuits is what keeps the active board an accurate picture of the work in front of you.

When a pursuit does advance into Proposal in Progress, Cairn starts the proposal for you. That's the next post.

Questions people ask

Can I add an opportunity by hand?

No. Every pursuit starts from a real notice or award: a result in Find Opportunities, an entry in the dashboard's Top Matches, or a candidate promoted from Recompete Radar. That's what lets the rest of the product read the source document, gather intelligence, and shred the solicitation later.

Do the NAICS, set-aside, and state filters apply to Grants.gov results?

No. Those structured filters only narrow SAM.gov contract results, because Grants.gov opportunities don't carry those fields. Grants are matched on your keywords alone, and your organization type on the company profile is what determines which grants you're eligible for.

What happens when I close a pursuit as Won, Lost, or No-bid?

It leaves the active board for the collapsed Closed strip at the end, stops counting against your plan's active-pursuit limit, and keeps its full activity log. Won outcomes are marked with a gold chip. To reopen a closed pursuit, move it back to an active stage; reopening counts it against the limit again.

How is the P-win calculated?

It's a deterministic weighted roll-up of the scores you enter on the bid/no-bid scorecard, using each criterion's weight. The same inputs always produce the same number, only criteria you actually scored count, and unscored criteria are excluded rather than treated as zero. An AI-suggested score is a draft until you accept it.

Open Company › Company Profile. The page has an auto-fill hero at the top, then four sections: Identity, Capabilities & targeting, SAM.gov API key, and, for grant applicants, Grants.gov submission.

Start with auto-fill from SAM.gov

Type your UEI into the box under Auto-fill from SAM.gov and click Auto-fill. Cairn calls SAM.gov and pulls your legal name, CAGE code, address, registration status, NAICS and PSC codes, and the set-aside certifications on your registration.

Two things we want you to know about how this works. Auto-fill only touches the fields SAM.gov returns, so anything you've typed into other fields stays put. And nothing is saved until you review the result and click the save button at the bottom, so you can run it, look, and back out.

Once SAM.gov data is in, a status strip shows whether your registration is Active (eligible to receive federal awards) and when it expires unless renewed. Both refresh each time you auto-fill, and the bell in the top band warns you as the expiry date approaches. Letting a SAM.gov registration lapse is the most avoidable way to lose an award, so we made sure the app nags about it.

Identity

Four identifiers live here. The UEI is the 12-character Unique Entity Identifier assigned in SAM.gov; it replaced the DUNS number in 2022, and the one you enter must match your registration exactly. The CAGE code is the 5-character Commercial and Government Entity code from that same registration, which agencies use to identify your business on contracts and payments; auto-fill sets it, so you'll rarely type it. The EIN is your 9-digit IRS tax ID, stored encrypted at rest and used only to match your entity. It never appears in searches or exports. And DBA names are trade names you operate under besides your legal name, comma-separated, so awards and opportunities recorded under those names still match you.

Capabilities and targeting

This section is what fit ranking reads. We'd give it more attention than any other part of the page.

NAICS codes classify what your business does; PSC codes classify what the government is buying. Both fields take a comma-separated list. The first NAICS code you list is flagged as your primary and shown as a gold chip. It's your headline classification for fit ranking, so put your most representative code first, not the one you happened to register first.

List the certifications you hold under Set-aside certifications: 8(a), WOSB, HUBZone, SDVOSB, and so on. Auto-fill sets these from your registration. Listing them is what lets fit ranking surface set-aside work you can actually bid, and it's also what lets Cairn flag a high-fit notice as Ineligible when your certifications bar you from it.

Agencies you name under Target agencies get extra weight in fit ranking, and Recompete Radar watches them for expiring awards. Comma-separated, and as specific as you can make them.

Under Contract size, give a minimum and maximum dollar band. Opportunities well below or well above it rank as weaker fits. Be honest here. A band that runs from nothing to everything tells the ranker nothing.

Pick your Organization type: it's your applicant type for federal grants, whether small business, nonprofit 501(c)(3), state or local government, tribal, higher education, and so on. Grants.gov restricts many funding opportunities to specific applicant types, so this one field decides which grants you're eligible for.

The last item in the section is the Recompete radar checkbox. Turn it on and Cairn runs a weekly sweep of expiring federal awards in your NAICS codes and target agencies, so you can position for the follow-on before the solicitation drops. It's part of the Professional plan.

The capability narrative

A plain-language description of what your company does, your past performance, and what sets you apart. The AI reads it both when scoring fit and when drafting proposals, so it's worth writing properly rather than pasting a tagline from your website.

If you'd rather not start from a blank box, Generate with AI drafts one for you. Click it and pick a source: Website (a URL, and we'll read the page), Document (a capability statement or brochure as PDF, Word, an image, or text, up to 10 MB), or Paste text. Click Generate narrative, then edit what comes back before you save. The source you provide isn't stored. AI drafting is included in the Professional plan; on Starter the button reads Generate with AI: upgrade and takes you to the subscription page.

There's an Additional context box below the narrative for anything else that should inform fit and the assistant. Unusual past performance, a niche you're moving into, a customer you'd rather not work for: this is the place.

Your SAM.gov API key

This section trips people up, so here's the whole story.

Auto-fill and opportunity search run against SAM.gov, and SAM.gov meters access with API keys. A personal key runs those calls on your own rate limits instead of a shared pool, which is why a key is required to save your profile. It's stored encrypted and never shown again; leave the field blank on later visits to keep the key you already saved.

To get one:

1. Go to SAM.gov and sign in. You'll need a Login.gov account, so create one if prompted.
2. Select your name in the top-right corner, then choose Account Details.
3. Scroll to the API Key section.
4. Click Request API Key and re-enter your password if asked.
5. Copy the generated key. It's shown only once, so store it somewhere safe.
6. Paste it into the API key field in Cairn and click Save changes.

So that setup works immediately, new accounts borrow a shared starter key for a short period. Once that period ends, opportunity search and saved-search alerts need a key of your own. Filling in your profile from SAM.gov keeps working either way, so you're never locked out of finishing the profile, and you'll see a reminder before starter access ends.

When the key expires

SAM.gov keys stop working roughly 90 days after they're created, and SAM.gov doesn't publish the exact date through its API. So we estimate the expiry when you save a key and remind you as it approaches: two weeks out, then again in the last few days. If SAM.gov showed you a different date, correct it in the Key expires field and the reminders follow the stored date.

An expired key is still sent to SAM.gov, because the date is only our estimate and the key may well still work. If it doesn't, you'll see errors from SAM.gov. Request a new key, paste it in, and save; saving a new key resets the expiry.

Save, and what happens next

The button at the bottom reads Create profile the first time and Save changes after that. On save, the "Finish your company profile" banner disappears across the app, and from then on new captures are scored against what you entered.

If you'll be filing grants through Cairn, the Grants.gov submission card at the bottom of the page is where the AOR certificate goes. That's an Enterprise feature and a topic of its own, so we cover it in part two of the proposal writer guide.

One habit worth forming: come back to this page when something real changes. A new certification, a new target agency, a narrative that finally says what you do now rather than three years ago. The ranker can only be as good as the profile it reads.

Questions people ask

What if my SAM.gov registration is about to expire?

Cairn shows your registration status and expiry date in the strip under the auto-fill hero, both refreshed each time you auto-fill, and the notification bell warns you as the date approaches. Renewing happens on SAM.gov itself; once you've renewed, run auto-fill again to pull the new status and date into your profile.

Does auto-fill overwrite what I've already typed?

No. Auto-fill only fills the fields SAM.gov returns, so your own edits to other fields stay untouched, and nothing is saved until you review the result and click Save changes. If you don't like what came back, leave the page without saving and your stored profile is unchanged.

Why is one of my NAICS codes shown in gold?

The first NAICS code you list is your primary code, and the gold chip marks it. It's your headline classification for fit ranking, so put your most representative code first. The order of the rest doesn't matter.

Do you keep the website or document I give Generate with AI?

No. The source you provide for the narrative, whether a URL, an uploaded document, or pasted text, is read to draft the narrative and isn't stored. What gets saved is the narrative text you review and keep, in your company profile.

Create your account

Sign-up is plan first. Go to cairnfed.ai/signup, fill in your name, email, and a password on the Create your account form, and click Continue. You'll land on Choose Your Plan. Pick Starter (it's free) and click Get Started. The account is only created once you've confirmed a plan, which is why there's no "confirm your email later" limbo.

Starter covers fit-ranked search and the pipeline. Professional adds the AI toolkit: the assistant, the proposal workspace, AI drafting, Recompete Radar, and team seats. You can start free and upgrade later without losing anything; a feature outside your plan shows as locked rather than hidden, so you can see what's there before paying for it. Current prices are on the home page.

After the plan step, Cairn sends you straight to the company profile. That's deliberate.

Fill in your company profile first

A brand-new account has a mostly empty dashboard, and the reason is always the same: nothing has been scored yet, because there's no profile to score it against. Fit scores, the dashboard's top matches, the weekly recompete sweep, and AI drafting all read from this one page. Until it's filled in, Cairn shows a banner that says Finish your company profile. on every page, with a Complete profile button. We put that banner there on purpose. It's the one nag in the product.

Here's the fast path:

1. Open Company › Company Profile.
2. Type your 12-character UEI into the box at the top and click Auto-fill. Cairn pulls your legal name, CAGE code, address, registration status, NAICS and PSC codes, and set-aside certifications straight from SAM.gov. Nothing is saved until you review it and click the save button, so you can look before you commit.
3. Add what SAM.gov doesn't know about you: your target agencies, your typical contract size, your organization type (this is what decides which grants you're eligible for), and a capability narrative that says what you actually do.
4. Paste in a SAM.gov API key and click Create profile.

That last step surprises people, so a word on it. The API key is required to save the profile, because auto-fill and opportunity search run against SAM.gov on your own rate limits rather than a shared pool. New accounts borrow a shared starter key for a short period so nothing blocks you on day one, and you'll see a reminder before that access ends. Requesting your own takes a couple of minutes on your SAM.gov account details page. The company profile post walks through every field and the key, including what happens when a key expires.

The capability narrative deserves more than a tagline. The AI reads it when scoring fit and when drafting proposals, so a plain paragraph about what you do, who you've done it for, and what sets you apart pays back everywhere. On Professional, Generate with AI will draft a first version from your website, a capability statement, or pasted text, and you edit from there.

Run your first search

Open Capture › Find Opportunities. One search covers both federal registries: contract opportunities from SAM.gov and grant funding opportunities from Grants.gov. Type a few keywords for the kind of work you do, keep both source pills on, and search.

Every result carries a fit score from 0 to 100 and a short Why it fits list. The score is how well the notice matches your profile: your NAICS and PSC codes, the set-aside certifications you hold, your target agencies, and your narrative. Treat it as triage rather than a verdict. It tells you what to read first, and clicking a title opens the official notice at its source so you're always reading the government's own posting.

Two things worth knowing on this page. First, the structured filters (NAICS, set-aside, state, due window) only narrow SAM.gov results, because Grants.gov opportunities don't carry those fields. Grants are matched on your keywords alone. Second, a high-fit result your certifications bar you from gets an explicit Ineligible flag instead of quietly disappearing. We'd rather show you a strong match you can't bid than have the score nudge you toward work you're not permitted to win.

When something looks promising, click Add to Pipeline. The button changes to Added to Pipeline and the opportunity becomes a pursuit.

Add a pursuit and look at the board

A new pursuit arrives in Identified, the first capture stage, and shows up on the board under Capture › Pipeline. The stages run left to right: Identified, Qualifying, Pursuing, Proposal in Progress, Submitted. Then a pursuit closes as Won, Lost, or No-bid.

Each card has an Advance button, and opening a card shows four tabs: Overview (the opportunity facts), Intelligence (AI-gathered details once you click Sync), Tools (the bid/no-bid scorecard and, later, the proposal and submission package), and Activity (every move and note, with a history behind it). Moving a card is a decision you're recording, not a drag on a whiteboard, which is why every move lands in the activity log.

We cover qualifying, the scorecard, and gate reviews in the pipeline post. For day one, adding one real pursuit is enough to make the rest of the product light up.

Read the dashboard

The dashboard answers one question: what deserves attention today. Four tiles across the top: Active Pursuits (everything not yet Won, Lost, or No-bid), Est. Pipeline Value (the estimated dollar value of every active pursuit, summed straight up and not weighted by win probability), Avg Fit Score, and Due This Month. Anything due within 14 days counts as critical and turns red across the whole app.

Below the tiles, Top Matches lists recently captured opportunities ranked against your profile, each with an Add button, and Upcoming Deadlines lists what's due on active pursuits. The date box on the right shows the federal fiscal year and quarter, because federal buying follows that calendar rather than the civilian one. If Top Matches is empty, go back to the profile: new captures are only scored once there's a profile to score them against.

Ask the assistant, and click the little icons

The floating button in the bottom-right corner opens the Cairn Assistant. It knows which page you're on, so "why is this ranked so low?" works without you explaining the context, and it can research opportunities or update pursuits for you. One guarantee we hold to: it never changes your data on its own. When it wants to add, update, or delete something, it pauses, lists each proposed action, and waits for your approval. The assistant is part of the Professional plan; on Starter the panel stays locked.

Small info icons sit next to fields and headings throughout the app. Click one for a short explanation of that specific thing. Each page also has a Page tour button in the breadcrumb strip that replays its guided walkthrough, and the full product documentation lives under Docs in the top menu once you're signed in.

That's the first hour. Profile, one search, one pursuit. Everything else in Cairn builds on those three.

Questions people ask

Is there a free plan?

Yes. Starter is free and covers fit-ranked search across SAM.gov and Grants.gov plus the capture pipeline. Professional adds the AI assistant, AI drafting, the proposal workspace, Recompete Radar, and team seats, and Enterprise is unlimited usage with priority support and automated Grants.gov submission. Current prices are always shown on the home page.

Do I need my own SAM.gov API key on day one?

No. New accounts borrow a shared starter key for a short period, so auto-fill and search work immediately. Once that period ends, opportunity search and saved-search alerts need a key of your own, which you request from the API Key section of your SAM.gov account details page. Filling in your profile from SAM.gov keeps working either way, and Cairn reminds you before starter access ends.

Can my teammates work in the same workspace?

Yes. Pursuits belong to your company rather than to one person, so everyone sees the same pipeline. Team collaboration is part of the Professional and Enterprise plans, and seats come with the plan: ten on Professional, unlimited on Enterprise. You add people from Company › Team, either by creating their account directly with Add user or by sending them a single-use invite token.

Is there a mobile app?

A Cairn companion app for iPhone is on its way to the App Store. It signs in with the same account, shows the same live pipeline, and carries your ranked matches and deadline alerts, and there is nothing extra to buy. Create your account on the web first; the app has no sign-up screen of its own.