The Books Are Always Behind
Here's the uncomfortable truth about startup accounting: by the time your books close, they're already fiction. You ran on Stripe for payments, Brex for cards, Gusto for payroll — tools that update by the second. But your accounting software? It sits there like a dusty filing cabinet, waiting for someone to export a CSV and manually categorize 4,000 transactions.
Puzzle looked at this mismatch and decided to fix the accounting layer instead of pretending it was fine. The company raised a $15 million Series A in February 2023, led by General Catalyst, with participation from FOG Ventures, XYZ Capital, Felicis, Sterling Road, Kapor Capital, and angels including Gokul Rajaram — the Meta board member whose investments span from fintech to AI. The round backed a simple thesis: if your financial tools speak in APIs, your accounting should too.
A Platform, Not a Spreadsheet Replacement
CEO and co-founder Sasha Orloff calls Puzzle "the first smart accounting platform designed for today's technology stack." That's a big claim, but the architecture backs it up. Puzzle connects directly to Stripe, Brex, Mercury, Ramp, Rippling, Gusto, and Plaid via their APIs. Transactions flow in automatically. Categorization happens in real time.
This isn't just auto-import — a feature many tools have added as a bandage. Puzzle was built from scratch as a real-time financial data platform wrapped in accounting software. The distinction matters. Legacy tools like QuickBooks and Xero were designed in an era where the source of truth was a paper receipt or a bank statement you downloaded weekly. Puzzle starts from the premise that the data is already streaming.
Co-founder David Liu, a former Stripe engineer, assembled a team with roots at Stripe, Square, and Brex. That pedigree shows up in the product's orientation: it thinks in events, not entries.
What You Actually See on the Dashboard
The output is where the architecture pays off. Puzzle generates detailed financial statements — profit and loss, balance sheet, cash flow, updated continuously rather than after a month-end close ritual. Startup metrics like true cash, burn rate, runway, and ARR sit front and center, computed from live data rather than a spreadsheet someone updates every other Tuesday.
There's a feature called "spotlight" that runs trend analysis across your financial picture. It surfaces anomalies and patterns you'd otherwise miss until a quarterly review. Spend analysis breaks down costs by team and category. Revenue tracking goes granular: by customer, by product, by account.
For multi-currency operations, and many startups now sell globally from day one, Puzzle handles that natively. This isn't an add-on bolted on later. The Synapse banking-as-a-service model showed that API-first financial infrastructure can serve diverse geographies; Puzzle takes the same principle upstream into the accounting layer.
The AI Layer and What Comes Next
The company has since evolved into what it calls "AI Native Accounting." Puzzle 2.0 launched with AI-driven accrual automation, automated revenue recognition, and what they brand as "Accuracy AI", essentially auditability built into the machine learning pipeline. The platform also introduced an API and MCP (Model Context Protocol) interface, positioning itself as agent-native infrastructure that other software can build on top of.
This is part of a broader pattern across technology innovation right now. The same wave that drove social technology forward, the same engineering momentum behind VR and AR platforms, is now reshaping how financial data gets processed and interpreted. Innovation in one domain rarely stays contained. Mental health technology researchers discovered this when they realized their funding models needed the same financial visibility that software startups demanded. A food and agriculture sustainable technology innovation programme managing grants across three continents faces the identical problem: your money moves through modern rails, but your books are still stuck in the spreadsheet era.
Puzzle's answer to that cross-sector need is the financial data platform embedded beneath the accounting interface.
The Fintech Inclusion Angle
There's a quieter story here about access. Historically, real-time financial visibility, the kind that lets you course-correct before a cash crunch becomes a crisis, was available only to companies that could hire a CFO and a controller and a finance team. Puzzle brings that capability to a two-person startup or a fintech serving underbanked markets.
The inclusion angle isn't marketing fluff. It's structural. When accounting software can read APIs, the barrier between a founder and their actual financial position drops from "hire an accountant and wait three weeks" to "open a dashboard." That shift compounds for founders in regions where finance talent is scarce or expensive, areas where the Latin American fintech boom has created urgent demand for tools that don't assume a full back-office.
The Investment Thesis Behind Puzzle
General Catalyst's lead on the round wasn't a bet on accounting software per se. The firm has a pattern of backing infrastructure that sits at the intersection of data pipelines and business operations. Puzzle fits that mold: it's a data platform that happens to produce financial statements.
The Series A will fund operations expansion and continued development. The investor list reads like a who's-who of startup ecosystem players, Village Global, OnDeck, a Y-Combinator alumni syndicate. Orloff and Liu's team has grown since the seed stage, pulling in engineers who want to work on problems that merge AI with financial infrastructure.
The governance and accountability question is real here too. As AI accountability frameworks mature across industries, Puzzle's approach of building auditability directly into its AI categorization, rather than bolting it on after regulators ask, positions it ahead of that curve.
What Puzzle Gets Right (And Where It's Still Early)
The product nails the integration problem. If you're a startup using modern fintech tools, connecting them to accounting without manual work is the highest-leverage automation available. Puzzle does this cleanly.
What remains unproven is whether the "financial data platform" positioning can pull Puzzle beyond the narrow band of venture-backed startups that use Stripe, Brex, and Mercury. Traditional small businesses don't have these integrations. They have a bank account, a shoebox of receipts, and a bookkeeper. Puzzle's answer, software plus a vetted accounting firm, acknowledges this, but it also means the company is competing with service businesses as much as with software products.
That's a reasonable expansion. It also means the company hasn't fully committed to being purely a data platform. Sometimes the answer is just: hire a good accountant and give them better tools.
The Bottom Line
Puzzle is building toward a future where accounting isn't a monthly chore but a continuous readout. The API-first design means the data pipeline is the product, the financial statements are a view on top. That's a genuinely different architecture from anything in the legacy accounting world.
The $15 million bought them a runway to prove that founders will pay for this, that the AI categorization holds up at scale, and that "real-time books" isn't just a sales phrase. The early signals, the team quality, the investor caliber, the integration depth, suggest they'll get that chance.
