The lab notebook is a financial document

What was broken

A deep-tech RF and microwave engineering startup needed a non-dilutive funding path from prototype to a scalable product.

What we did

CapEasy shaped the seed-fund application around permitted activities — phased prototype-to-market-entry milestones and a compliant budget. We built the supporting financial model and aligned the proposal to the programme’s eligibility rules.

Where it landed

Approved for a seed grant through an incubator-backed programme; disbursement follows diligence. A seed-fund approval is the incubator committee’s sanction; actual disbursement follows the programme’s due-diligence and compliance milestones and is not guaranteed.

RF hardware has two federal doors, and they read your file differently

NSF does not solicit specific technologies — its SBIR/STTR Phase I program (solicitation NSF 26-510) funds "across nearly all technology areas and market sectors" and caps Phase I at $305,000 for a 6-to-18-month project, so an RF or microwave engineering team competes on technical merit and commercial potential rather than fitting a pre-published topic. DoD runs the opposite model: individual components publish their own numbered topics, and RF/microwave hardware shows up regularly inside them — Army SBIR runs topics through its own component sites and the government-wide BAA at sbir.gov, and Navy and Air Force components have run open RF-front-end and RF-component topics on their own release cycles. The topic list turns over every solicitation cycle, so the specific open topic and its deadline has to be checked on sbir.gov’s live topic search before a proposal gets written, not assumed from a prior cycle.

The engagement’s discipline was matching the proposal to what the scheme actually permitted — prototype-to-market-entry milestones, nothing outside the funded categories. The same discipline decides which federal door fits: NSF for an open technical bet on a Phase I budget cap, a DoD component topic for a mission-specific RF problem with its own phase structure and end user named in the solicitation.

The lab notebook is the source document — for the grant and for the tax credit

RF and microwave R&D produces a specific kind of evidence: dated test logs, S-parameter sweeps, noise-figure measurements, a record of what was tried and what changed between one prototype revision and the next. That record is not just engineering hygiene — it is the primary substantiation the IRS looks for when a company claims the Section 41 research credit. The IRS’s own Audit Techniques Guide states that "contemporaneous books and records should form the basis of the examination," and that qualification is judged on what an engineer actually did during a specific period — payroll records, time allocation between qualified and non-qualified work, and project documentation — "should not be based solely on job descriptions or titles."

A funded grant milestone and a qualified research expense are, in practice, read from the same underlying file: a dated notebook entry or test log tied to a specific engineer, a specific hypothesis, and a specific result. A team that keeps that record as it happens — not reconstructed from memory during a filing deadline or a diligence window — can point a grant reviewer and, separately, an R&D credit preparer at the same evidence. Kept casually across engineers’ personal notebooks and lab computers, it becomes weeks of reconstruction the second either reviewer asks for it. CapEasy’s part is exactly what the engagement did: turning that engineering record into a financial file that ties to a milestone and a ledger; the credit calculation and its filing position run through the partner CPA firms we work with across 15 US states.

The budget justification names a person and a rate, not a round number

NSF’s PAPPG requires senior and key personnel to be budgeted by name, with a stated calendar-month commitment — a Phase I proposal budgets the PI for at least one calendar month of effort per six months of the project — and a documented rate behind each line, the same standard the sibling engagement applied when it mapped prototype and engineering costs into the scheme’s permitted budget categories with no marketing or commercialisation spend folded in. A budget line that reads "engineering, $180,000" without a personnel table behind it fails the same test on either side of the Pacific: a reviewer cannot verify a number nobody built.

The engineering-specific version of this problem is test-and-measurement equipment, not a bill of materials. RF work is claimed on instrument readings — a vector network analyser’s S-parameter trace, a spectrum analyser’s noise floor — and NIST’s own policy on metrological traceability is explicit that a result is only traceable when it can be related to a reference standard "through a documented unbroken chain of calibrations." An uncalibrated or undocumented instrument does not just risk a bad measurement; it undermines the record a grant milestone claim or a research-credit test log depends on to prove the work happened as described. Calibration certificates, due dates and instrument logs belong in the same file as the notebook entries they support.