FedGrade watches SAM.gov solicitations, DHS acquisition forecasts, Grants.gov, and federal award feeds, and tells you which ones belong in your pipeline.
Then it goes further: it drafts the proposal volumes, builds invoices from your approved timesheets, assembles submission packages, and tracks every compliance deadline. You review and approve everything before it moves.
FedGrade drafts, assembles, and tracks — you stay the signature on everything that leaves the building.
Biddable-only discovery across SAM.gov notices, DHS acquisition forecasts, Grants.gov, and a SAM.gov award-notice archive going back to FY2002. Open notices by default, no expired ones — just what belongs in your pipeline.
Opportunity finder — open notices only, with agency, pWin, stage and due in every row
Capture assessments, win probability, and teaming intelligence, plus a daily brief tuned to your NAICS codes and agencies — items link to their source record when one resolves, and are marked unsourced when it doesn't.
Pipeline — stage, gate decision, deadline and value per pursuit, rolled up above
DCAA-shaped timesheets, auto-drafted invoices for T&M, CPFF, and FFP contracts, and payment tracking with prompt-payment clocks. You authorize before anything is issued.
Submission packages assembled to Section-L order with signed receipts — plus the compliance layer: ISR/SSR, CPARS prep, SAM renewal, subcontract flowdowns, and filing deadlines.
Every AI tool in government contracting now says its answers are cited. Ask the next question: who checks the citation — and can you check it yourself? FedGrade's answer is a mechanism, not an adjective. A confident fabrication in anything a contracting officer reads is a bid risk, so the pipeline is built to make fabrication detectable by you.
Each claim in the FedGrade Brief is checked against its cited source: the quoted excerpt must be a character-for-character substring of the retrieved document. A draft that fails the check is redrafted, and if it still fails, the failure is shown on the Brief — it says there wasn't enough evidence instead of guessing.
When a Brief needs a second drafting pass, a second model also checks — when the checker is available — that each quoted passage actually supports the sentence it's attached to, and drops the ones it doesn't.
Generated export files carry an Ed25519-signed provenance manifest: a hash of the exact file, what it was built from, and how. Change the file and the hash no longer matches. Signing of individual Brief claims is not live yet.
Brief items link to the source record they cite where one resolves, and are marked unsourced where one doesn't, so your contracts lead or your CPA can open the evidence instead of taking our word for it. A public verification page is on the roadmap.
We're opening seats gradually ahead of launch. Tell us about your agency focus and pipeline, and we'll reach out as seats open.
Brief claims are validated against their cited sources, and failures are shown, not hidden; claims that needed a second drafting pass are also checked by a second model when the checker is available. SOC 2 Type II and FedRAMP Moderate are roadmap targets, stated as targets. Data stored in US regions, AES-256 at rest, TLS 1.2+ in transit. FedGrade is not affiliated with, endorsed by, or acting on behalf of any government agency.