Boast.ai Is Selling Months, Not Magic Credits

In February 2021, Boast co-founder Lloyed Lobo told Crunchbase the sentence that still explains the company better than any homepage slogan: companies burn R&D all year, then wait 16 months or more for the financial benefit because credits hitch a ride on tax filing and government processing. That is not a branding problem. It is a cash-timing problem wearing a tax costume.

Boast (legal entity Boast Capital LP, product site boast.ai) sells against that lag. The brochure leads with AI and “innovation capital.” The purchasable product is closer to: continuous evidence so the claim survives review, plus an optional advance so you are not financing last year’s engineers with this year’s equity raise. If you evaluate Boast as “an R&D credit explainer with a feature list,” you will buy the wrong thing—or skip the right one for the wrong reason.

The credit is public policy. The wait is your P&L.

United States filers chase federal research credit under Internal Revenue Code Section 41 (and a patchwork of state programs). Canadian companies chase SR&ED at the CRA, plus provincial layers such as IDMTC and Quebec’s CDAE / CDAE-IA. None of those statutes promise liquidity on the day you ship a prototype.

So founders invent workarounds:

• Ignore the credit until Series B finance “has bandwidth.” Leave refundable cash on the table for years.
• Hire a year-end scramble. Someone interviews engineers in March about experiments from last June. Narratives get written. Evidence gets reconstructed. Audit risk climbs with every reconstructed week.
• Borrow against the hope. A bank or specialty lender fronts something; you pay interest while the claim is still a PDF in progress.

Boast’s commercial bet is that the middle path—wire GitHub / Jira / payroll / accounting into a claim system, keep specialists on qualification, and optionally draw against expected credits—beats all three. Whether that bet clears for you is almost entirely a net-proceeds and risk question, not a vibe question about “AI tax.”

Strip the brochure: three layers, one clock

Boast markets a hybrid, not a calculator:

1. Software that ingests project, payroll, and accounting signals to flag candidate activities and Qualified Research Expenses. Product and solution pages name stacks such as GitHub, Jira, payroll/HR (Gusto, Rippling, ADP, Workday), and accounting (QuickBooks, NetSuite). The point is contemporaneous evidence—commits, tickets, time, costs—not a heroic retrospective interview.
2. In-house specialists who still decide what is technologically uncertain, what counts as experimentation, and what belongs in a defensible file. CEO Imad Jebara’s March 2026 commentary around the benchmark report is blunt on this: AI accelerates collection; it does not replace judgment on technical uncertainty. Audit-facing packaging includes offerings such as AuditShield.
3. QuickFund — marketed especially on the Canadian SR&ED side — advances up to roughly 75% of expected credit value, with draw talk in the $25K–$1M quarterly band, funding often described in about two weeks, and access pitched as up to 12 months earlier than waiting for the refund. Interest accrues on what you draw; principal typically settles when the government money lands. Exact terms are engagement-specific.

Geography is narrow on purpose: U.S. and Canada, with localized sites (/en-us, /en-ca, /fr-ca). Cross-border teams are the quiet differentiator versus Canada-only or U.S.-only tools. In December 2025, Boast also launched a dedicated tech-enabled path for Quebec’s CDAE-IA transition effective 1 January 2026—another reminder that “R&D credit” is not one product SKU.

Security copy includes role-based access, MFA, and SOC 2—non-optional if you will actually connect payroll and source control.

The $100M facility is not a Series A with better PR

Public funding milestones worth keeping straight:

• Operating since 2011 (company materials; some early press said 2012—prefer Boast’s own founding claim).
• December 2020 Series A: $23M USD led by Radian Capital (company PR; some Canadian coverage rounded near $30M CAD). Near 1,000 customers at announcement; offices then listed across San Francisco, Toronto, Vancouver, Calgary.
• February 2021: $100M credit facility from Brevet Capital Management to fund advances against R&D credits and subsidies (Crunchbase News; company release). Debt / specialty finance—not equity.
• Leadership: Alok Tyagi as CEO in 2022; Imad Jebara named CEO in March 2024 (ex-Diligent / Galvanize customer-operations leadership). Still listed as CEO in mid-2026 materials.

Crunchbase-style tallies that casually say “$123 million raised” smash equity and the Brevet line into one vanity number. Say $23M Series A plus a $100M facility or you are misreading the balance sheet story. The facility exists because the product thesis is front the months. Without that thesis, Boast is just another claim shop with better integrations.

Distrust the cohort just enough

March 2026: Boast’s 2026 R&D Tax Credit Benchmark Report (GlobeNewswire, 26 March 2026) published anonymized platform stats for the 2024 season: 6,907 claims, 2,298 companies, $3.6B+ tracked R&D spend, $900M+ credits secured. Average claim $768,233 (245% up since 2018, per Boast). Software/internet 80.6% of claims and lowest sector audit rate in-set (5.28%). U.S./Canada mix about 54.3% / 45.7%. Overall audit rate 6.59%; 93% of claims passing without review in their data. 90% SMB. Multi-jurisdiction strategies tied to 30–50% higher recoveries among “sophisticated” claimants in the release.

Homepage claims still lean hard: 2,000+ businesses since 2011; more than 98% of Boast claims delivered without an audit; when CRA/IRS contact happens, recovery of up to 95% of the estimated claim with Boast representing the client. About-page cumulative “innovation capital” has drifted across copies ($625M, $675M+, $900M+ depending on which page and which season you open)—prefer dated citations.

Read the report as Boast’s customer cohort, not CRA/IRS population truth. Useful for who shows up on the platform (software-heavy, SMB-skewed, often multi-program). Useless as a personal audit-odds guarantee. Vendor comparison tables that put client effort near ~5 hours versus 20+ or 40+ elsewhere are marketing benchmarks, not your SLA.

Fee math beats feature checklists

Third-party buyer guides in this category usually put contingency-style fees around 15–25% of credit value. Boast does not publish a public rate card; expect a sales-quoted, success-oriented model. QuickFund interest stacks on top of whatever success fee you already owe.

Run the ugly arithmetic before the demo:

• Tiny first-year claims: percentage fees + advance interest can erase the win. DIY or a boutique retainer may be cleaner.
• Material claims with messy engineering history: reconstruction cost (engineer hours, weak files, audit exposure) is the hidden line item Boast is trying to compress.
• Cross-border U.S.–Canada or Quebec CDAE-IA + SR&ED stacks: vendor fragmentation has its own tax—two processes, two narrative styles, two audit postures.

Competitors exist and are not cardboard villains: Big Four for complex multi-entity work; boutiques for deep local program knowledge; Canada-focused software such as GrantOps on different fee shapes; U.S. tech-led tools such as TaxRobot on speed and contingency packaging. Boast’s sharper edge, when it has one, is cross-border coverage plus mandatory specialist review plus the advance product. Do not buy “AI” as the differentiator. Buy or reject the months + defensibility package.

Napkin math: claim × fee × QuickFund interest

Boast will not publish a rate card. Your CFO still needs a number before anyone connects payroll. Build it from three inputs—expected claim, success fee, advance interest—then swap in the actual sales quote.

Worked example (illustrative only; not a Boast quote):

1. Expected credit / refund: $200,000 (a plausible mid-size SR&ED or Section 41 outcome—not Boast's published average).
2. Success fee: mid of the common buyer-guide 15–25% band → 20% → $40,000. After fee, if you simply wait for the government: $160,000.
3. QuickFund draw: Boast markets up to 75% of expected credit → $150,000 cash in roughly the two-week window they advertise, not the 12–16 month lag Lobo described.
4. Interest on the draw: specialty SR&ED advances in the open market commonly price around 1–3% per month on amounts drawn; use 1.5%/month as a mid placeholder for 10 months until settlement → $150,000 × 1.5% × 10 = $22,500. (Real offers may quote APR, minimum interest periods, legal/disbursement fees; Boast's exact rate is engagement-specific.)
5. Net when the refund lands: $200,000 − $40,000 fee − $22,500 interest = $137,500. You did not "make" $137,500 versus waiting—you made $150,000 of liquidity ~a year early, and paid ~$22,500 plus process overhead for that clock shift.

Run the same stack on a $40,000 first-year claim: 20% fee is $8,000; 75% draw is $30,000; ten months at 1.5%/month is $4,500. Net after fee and interest ≈ $27,500, with payroll and source control now wired to a vendor. That is usually where DIY or a boutique retainer wins—and where "AI tax" demos waste a quarter.

Stress the model before the demo, not after: what if CRA/IRS trims the claim 20%—who covers the advance shortfall? What is the minimum interest period if the refund arrives in four months instead of ten? Are Form 6765 / Quebec CDAE-IA programs even in the quoted scope? Those answers belong in the engagement letter. If the napkin only clears because you assume zero audit risk and a homepage interest vibe, you are still buying decoration.

Boast is a mature North American operator: 2011 services roots, AI-assisted capture, humans on technical uncertainty, and a specialty facility that productizes impatience. The statute did not change. The months did. Price the months with arithmetic. The website is everything around that fight.