Federal Capability Coverage · Gap Analysis

Before you commit a team to a pursuit, can they actually cover what it calls for — or are you guessing?

Load your own team, a competitor’s, or one you’re assembling — up to five companies — against a specific contract pursuit or a buying office, and get a capability-area coverage read: covered, partial, gap, or honestly unknown. Holding the vehicle isn’t the same as having done the work, and this tool never lets it count as if it were.

See it work ↓

Included with Growth and Enterprise accounts. There are two kinds of number on this page and we label every one: the corpus figures just below are real, measured against the platform’s own published figures. The product screen further down is the real tool, run over a fictional team and a fictional pursuit that’s safe to publish.

What Every Verdict Is Checked Against

3,047,746 demonstrated capabilities, read off 1,011,617 federal awards — the same evidence every coverage cell on this page is checked against.

A covered cell isn’t the incumbent’s word, and it isn’t “they hold the vehicle.” It’s a dollar ratio against work the government’s own record — or, where a scope document has been read, the government’s own requirement — shows a company has actually performed. Measured 4 September 2026. Capabilities and requirements are read out of the government’s own documents with AI assistance (Claude, by Anthropic); every coverage percentage, count and dollar figure is then computed from the federal record.

3,047,746
Demonstrated capabilities on file
1,011,617
Federal awards behind them
55,690
Companies profiled
CoveredPartialGapUnknown
How A Cell Gets Its Color
Coverage is a dollar ratio against what the requirement — or capability area — was actually worth: under 50% is a gap, 50–99% is partial, 100% or more is covered.
Evidence is graded, and the strongest claim wins: a requirement read out of the government’s own scope document outranks a capability inferred from an award description, and inherited vehicle scope never counts toward a cell at all — and every run declares the depth of both sides, the requirements it read and the profile it judged them against.
A vehicle’s own inherited scope evidence is shown as support, never counted toward a coverage cell — it applies identically to every holder of that vehicle, so it can’t be what tells one company apart from another.
A fourth state, Unknown, exists specifically for when our own visibility isn’t deep enough to call it — never colored red for a limit that’s ours, not the company’s.
What A Red Cell Doesn’t Mean
Not “this company can’t do the work.” It means the visible federal record doesn’t show it as a prime, under a code the extraction caught, inside the visible window.
Not a company’s whole nationwide portfolio. The figures shown are the pursuit-relevant evidence — not “how big is this company, period.”
Not a fit score or a recommendation. Coverage percentages and counts are shown; nobody is told who to pick.
Not final. As more scope documents get read, a row can honestly move from Unknown to a real color — that’s the mechanism working, not a number moving under you.

The Workflow

Build the team. Name the target. Read the matrix.

Every step below runs on the exact same evidence tables the rest of the platform reads — nothing here is a simplified demo path. It’s one screen, not three: the walkthrough is told as three steps because that’s the order you’d actually work it, not because there are three different pages.

1Build The Team
Add up to five companies — yours, a competitor’s known team, or a mix you’re assembling. It works pointed either direction: plan your own pursuit, or see what the federal record says about a rival’s.
2Name The Target
Search for the specific solicitation or program you’re chasing, or a customer office you sell to. From any contract’s own profile, the target arrives pre-filled — one click, not a re-search.
3Read The Matrix
A capability-area-by-capability-area verdict, plus a Teaming Options list ranked by how many open gap rows a candidate covers — or, when the whole frame comes back Unknown, how many of those unknown areas it demonstrably covers instead. A plain count, never a fit score.
intelligent.win/intelligence/gap-finder

The real tool, top to bottom: a team row against a specific contract pursuit, a stat band stating how many capability areas are covered and how much of the contract's dollar value is accounted for, then the capability-area matrix itself — each row colored Gap, Partial or Covered, showing what the work called for against what the record shows. Below the matrix, Teaming Options ranks candidates by how many open gap rows they cover — or, when the frame is entirely Unknown, how many of those unknown areas they demonstrably cover instead — weighted by dollars, split into revenue bands (Large, Mid-Cap, Small, Challenger) so a $250M prime and a $10M small business are never compared as if they were the same kind of candidate. At the very bottom, the solicitation's own extracted requirements, shown in full wherever the government's scope document has been read.

In this image: companies are fictional and dollar figures are illustrative. Agencies, buying offices and codes are the real federal record; any contract number shown is altered so it resolves to nothing.

Showing the top of the page — full capture 5,838px tall, expand for all of it.

What You Actually Get

Everything Gap Analysis shows — and what it deliberately refuses to.

Coverage you can check yourself. Never a verdict on who should win.

In The Product
Up to five companies loaded as one team — yours, a competitor’s, or a mix — against a specific contract pursuit or a buying office
A capability-area coverage matrix: Covered, Partial, Gap, or honestly Unknown — a dollar ratio, never a score
Where the government’s own scope document — a PWS, SOW, or SOO — has been read, its real requirements shown in full, not just the capability-area buckets they’re extracted into
Teaming Options ranked by how many open gap rows a candidate covers — or, when a frame comes back entirely Unknown, how many of those unknown areas it demonstrably covers instead — weighted by dollars, split by revenue band so scale is never confused with fit
Vehicle position shown as a plain fact on every row — a vehicle’s own inherited scope evidence is disclosed as support, never counted toward a coverage cell
Deliberately Not Here
A fit score, a match percentage, or a recommendation of who should win — the evidence is assembled here; the team decision is still yours
A red Gap cell where our own visibility is too shallow to judge — that’s Unknown, a fourth state, never colored as if it were a finding
Coverage counted from a company holding the right vehicle alone — holding it isn’t the same as having done the work, and this tool never lets it count as if it were
A candidate’s whole nationwide portfolio standing in for pursuit-specific fit — ranking reads the evidence that actually matches this pursuit
Any claim that a company is or isn’t a small business — SAM’s own on-file certification is shown, never our verdict

Stop finding out your team can’t cover it after you’ve already committed.

Load your team against the pursuit before you write a single line of the proposal — and see, capability area by capability area, exactly what’s covered, what’s open, and who else could close it.

Or see how the whole platform works first →

Common Questions

Four things people ask before they trust a coverage read like this.

What does Gap Analysis actually do?

Gap Analysis measures demonstrated capability evidence against a specific contract or buying office, capability area by capability area — and it works just as well pointed the other way: load a competitor's known team and see what the federal record says about them. Coverage is a dollar ratio, not a score. It shows what the federal record supports and stops there — no fit scores, no recommendations. You decide who to call.

How do I get from a contract to a ranked list of teammates?

Load up to five companies as your team, then point the tool at a specific contract pursuit or a buying office and run it. You get back a capability-area coverage matrix — covered, partial, gap, or unknown — plus a ranked Teaming Options list of companies that could fill what your team doesn't cover. From any contract's own profile, the tool is one click away with the target already filled in.

What do the red, yellow, and green cells mean — and what don't they mean?

They're a dollar ratio against the requirement's own recorded value: under 50% is a gap (red), 50-99% is partial (yellow), 100% or more is covered (green) — plus a fourth, grey state, Unknown. Two things the colors do not mean. First, a red row doesn't mean your team can't do the work — it means the visible federal record doesn't show that work as a prime, under a code the extraction caught, inside the visible window. Second, an Incumbent or candidate figure is that company's whole portfolio-wide federal record, not what they've delivered on this specific contract. Sanity-check any hard gap against what you actually know before you act on it.

Gap Analysis marked a row "Unknown" instead of red. Why?

Because on that pursuit we're reading the government's own scope document — a detailed, requirement-by-requirement list — while the company we're checking is only visible to us through award descriptions, which are much coarser. Calling that a gap would be us asserting something we can't actually see. So the row goes grey and says Unknown, and it's excluded from the gap count. It is not a small gap and it is not a pass: it means our visibility isn't deep enough to judge that row yet. An Unknown row is not excluded from Teaming Options, though — if a whole screen comes back mostly grey, the tool still surfaces candidates, ranked against those unknown capability areas rather than confirmed gaps and labeled as such, because only your team knows whether you actually cover that work. This state got dramatically rarer in late August 2026 — the number of companies we can see deeply enough to call a real gap on grew by more than thirty times as we read more scope documents, so a screen that came back mostly grey before then is worth re-running. The same row can honestly change from Unknown to red later, and that's the mechanism working, not a number moving under you.