Most grant software stops at a searchable list. ISA Research runs the whole race: it finds live federal opportunities, scores your real fit honestly, drafts agency-aware proposals with you, builds a Uniform-Guidance-correct budget, tracks every submission requirement as a checkmark — and prints the finished application package.
Seven stages, one workspace. Every stage feeds the next — nothing is retyped, nothing is lost between tools.
Search the live federal opportunity feed (the public Grants.gov API) without leaving the app — agency, program, deadline, and award size at a glance, with your pipeline one click away.
The Forge Match engine scores every opportunity against your organization's real capability profile — six weighted components covering topical coverage, maturity, eligibility, award size, runway to deadline, and application burden.
A seven-criterion go / no-go rubric — mission fit, competitiveness, capacity, financials, timeline, relationships, burden — produces an evidence-cited recommendation. Conditional-go and no-go are real answers, not failure states.
The proposal workspace opens with agency-aware section templates (NSF, NIH, U.S. Dept. of Education, and a generic federal set). AI scaffolds each narrative section from your capability profile; you edit, it never submits anything on its own.
A budget builder that knows the Uniform Guidance (2 CFR 200): modified total direct costs, indirect rates and the de-minimis option, and dollar thresholds resolved by award vintage. It lints your budget before a federal reviewer does.
AI reads the funding notice and turns it into a requirements checklist — every form, attachment, and certification as a trackable checkmark with an audit trail. On award, a compliance calendar generates the RPPR, FFR, and effort-certification deadlines for the whole period of performance.
One click produces the submission package: cover page, table of contents, narrative sections, multi-year budget with rollups, and the requirements checklist — print-perfect, and stamped DRAFT until every required checkmark is complete.
Built as a module of the Luca AI Express platform — same account, same design language, same data fabric as the rest of the Luca family.
Search current federal funding opportunities in-app via the public Grants.gov API — no scraping, no stale mirrors — and save them straight into your pipeline.
Describe what your organization actually does once; the match engine scores every opportunity against it with IDF-weighted term coverage and graded — never binary — eligibility.
Seven weighted criteria, an eligibility gate, and thresholds that produce go, conditional-go, or no-go — each with per-criterion rationale and cited evidence you can defend to a board.
Sections with statuses and a state machine from draft to submitted; AI scaffolds each section from your profile and the opportunity, and every word stays yours to edit.
Multi-year budget lines with justifications, automatic MTDC and indirect rollups, and lint rules for the 2 CFR 200 thresholds that actually apply to your award's vintage.
Paste or attach the funding notice; AI extracts the submission requirements into a deduplicated checklist with owner and timestamp on every completed item.
From the award start date, the platform generates the reporting calendar — progress reports, federal financial reports, effort certifications, closeout — so nothing surfaces as a surprise.
A print-first application document — cover, contents, narratives, budget tables, checklist — rendered from live data, ready for File → Save as PDF.
A living research library — models, hardware, methods, datasets — plus PhD-style research notes with tags and cross-references, feeding the same workspace your proposals live in.
Forge Match is a versioned, deterministic engine — the same inputs always produce the same score, and every score decomposes into its components.
Each opportunity is scored 0–100 against your capability profile:
Strong fit — commit resources and draft.
Fundable with named gaps — partner, pilot, or descope first.
An honest no — with the exact reasons, so next cycle you can change them.
ISA Research was built under a strict data-isolation policy — sharing is an explicit act by the owner, never a default.
Every note, proposal, capability, and score belongs to one user. There is no org-wide pool, no ambient visibility, no leak path between accounts — isolation is enforced in every query, not in the UI.
Owners grant named collaborators exactly three possible rights — read, write, change. Delete is structurally ungrantable: the system has no way to give it away. Only the owner can revoke.
Every AI feature — NOFO parsing, section scaffolds, deep pre-submission review, match screening — runs on frontier-class Claude models through the platform's governed AI bridge.
The platform will tell you no. Scores that flatter cost you staff-months on unwinnable applications; scores you can defend build a funding strategy. We chose the second.
Federal education programs, ESSA evidence tiers, and EDGAR-formatted applications — with the compliance calendar your business office will thank you for.
NSF and NIH template sets, effort certification tracking, and honest competitiveness scoring across a whole opportunity pipeline.
Capability-matched discovery so a small development team spends its hours only on opportunities it can actually win.
A research corpus, PhD-style notes, and a proposal factory in one workspace — from literature to funded project.
ISA Research is rolling out in cohorts. Leave your email and we'll reach out with access details and onboarding for your organization.