What a government funding committee actually reads in your financial file

What was broken

A recognised deep-tech founder needed non-dilutive seed capital to move from a validated prototype toward a market-ready product — and had to present an application to a public seed fund that would clear an incubator selection committee.

What we did

CapEasy structured the proposal end to end: eligibility, the case for a convertible-debenture (scaling) instrument over a grant, and a utilisation plan mapped only to scheme-permitted heads. We built the milestone-linked tranche plan and the financial model the committee assesses, and shaped the narrative so the technical work read clearly to a non-specialist panel.

Where it landed

The application was approved for a convertible debenture through the incubator’s seed fund. A seed fund approval is the incubator committee’s sanction; actual disbursement follows the scheme’s due-diligence and compliance milestones and is not guaranteed.

The US programs that fund a scientific-software company, by name

For a deep-tech or scientific-software startup, "government funding" in the US is not one program — it is eleven federal agencies running Small Business Innovation Research (SBIR) and Small Business Technology Transfer (STTR) solicitations, each with its own topic areas, review criteria, and financial documentation expectations. The agencies most relevant to scientific software and instrumentation are NSF (general-purpose deep tech, plus a dedicated pilot for next-generation instrumentation and experimental platforms), NIH (software and platforms for biomedical or health research), DOE (computational and energy-science tools), and the Department of Defense — most visibly through the Air Force’s AFWERX, which runs its own SBIR/STTR track and a Direct-to-Phase-II path for companies with mature technology.

These are not interchangeable pots of money with a single form. NSF SBIR/STTR awards are not government contracts — NSF does not use the program to procure goods or services, so the award behaves more like the grant most founders picture. DoD normally extends SBIR/STTR as a contract, which comes with a results-based, deliverable-driven structure closer to a commercial engagement. Reading which structure your target agency uses, before you build the budget, changes what your books need to prove.

The financial file a reviewer actually opens

A federal SBIR/STTR reviewer is not grading a pitch deck. Award ceilings are set centrally — the standard SBA guideline is up to $323,090 for Phase I and up to roughly $2.15 million for Phase II, unless an agency has an SBA-approved higher ceiling — and NSF’s current Phase I award is up to $305,000 covering direct costs, indirect costs, and fee together. Inside that number the reviewer reads a budget justification: every direct-cost line tied to a specific task in the work plan, an indirect (overhead) rate that reflects your company’s own costs — not a borrowed or invented rate — and, for agencies like DOE, a minimum weekly time commitment from the named Principal Investigator so the PI does not read as window dressing on someone else’s project.

The same reviewer, or the contracting officer behind them, expects your accounting system to be able to produce that justification on demand. SBIR guidance points applicants toward a FAR-compliant accounting system meeting SF-1408 criteria — one that segregates direct from indirect costs, tracks labor by project, and excludes unallowable costs from the start rather than backing them out later. A file assembled after the fact, from spreadsheets that do not tie to the ledger, is the federal-funding equivalent of the undocumented promises an incubator committee has learned to distrust.

Grant, contract or convertible instrument — know which one sits on your balance sheet

The LabCanvas engagement is instructive precisely because the fund chose a convertible debenture over an outright grant — a structure with its own accounting consequence, and one worth understanding before a US founder assumes every public-sector dollar behaves the same way. A pure SBIR/STTR award is non-dilutive: the government takes no equity and the funds are typically recognised as grant or contract revenue against the work performed, not as debt. A convertible instrument from a fund, accelerator, or state-backed vehicle is different — it is a liability (or, depending on its terms, a compound instrument) that sits on the books from the day it is signed until it converts or is repaid, the same way a SAFE does.

FASB simplified this corner of GAAP with ASU 2020-06, which reduced the number of accounting models for convertible debt and removed the requirement to split out beneficial-conversion and cash-conversion features from the host instrument in most cases — more convertible debt is now carried as a single liability at amortised cost rather than split across multiple lines. That simplification does not remove the discipline it assumes: a schedule of every convertible instrument — amount, date, conversion terms, and the signed agreement filed against it — kept current from signature, not reconstructed when a future investor or agency auditor asks for it.

The milestone-linked tranche plan that got the LabCanvas application past committee transfers directly here: whether the instrument is a grant, a contract, or a convertible note, US programs increasingly release money against demonstrated milestones rather than a lump sum. Cost tracking that can show, task by task, what each tranche funded and what it produced is the artefact that both a US federal reviewer and a convertible-note holder are ultimately reading for.