Inscope.

There is a genre of corporate suffering that never makes the news because the people enduring it are too busy tying out a cash flow statement to complain. Inscope is a bet that this suffering is, in fact, a market.

Here is a thing that is true and slightly absurd: some of the most important documents in capitalism - the financial statements that tell investors, lenders and auditors whether a company is a going concern or a smoking crater - are still assembled by hand. Not literally by hand, but close. A controller exports numbers from an ERP system, pastes them into Excel, formats them into Word, emails the draft to a colleague, receives it back with tracked changes, discovers that a number moved upstream and now nothing ties, and starts over. This happens every quarter. It happens at companies you have heard of. It is the reason a certain kind of finance person develops a thousand-yard stare in mid-February.

Inscope, a San Francisco company founded in 2023, has decided that this is the problem worth spending a career on. The pitch is unglamorous in the way that the best B2B pitches are unglamorous: it automates the preparation of financial statements. Footnotes, disclosures, the dreaded cash flow statement, the formatting nobody agrees on. According to the company, it automates more than 60% of the work. According to its own homepage, the goal is financial reporting "so easy, you might think it's fun," which is the kind of tagline you can only write if you have personally not found it fun.

The founders lived the problem, which is the whole point

The two people at the center of Inscope are Mary Antony and Kelsey Gootnick, and their resumes are the argument. Both are CPAs. Both ran accounting and finance functions at high-growth companies - the names that come up include Flexport, Dropbox, Miro, Autodesk and Yelp. They met roughly seven years before starting the company, at Flexport, where Gootnick was the controller and Antony was the assistant controller, which is to say they spent their formative professional years at adjacent desks doing exactly the work they now want to abolish.

This matters more than it might for a consumer app. Financial reporting is not a domain you can vibe your way into. It is full of edge cases, materiality thresholds, judgment calls and the specific institutional knowledge of why this footnote is worded that way. Norwest partner Sean Jacobsohn put it about as bluntly as a venture investor puts anything: "Few founders possess the specific expertise required to reinvent financial reporting technology." The subtext is that a lot of people have tried to build accounting software without knowing accounting, and it shows. Antony and Gootnick were later joined by co-founder and CTO Jared Tibshraeny, which is the part where the domain experts acquire the person who can make the LLMs behave.

What it actually does

The product reads from your ERP - Oracle NetSuite is the one that comes up most often - and uses large language models to generate GAAP-compliant statements and the surrounding apparatus of footnotes and disclosures. It checks the math. It enforces consistency, so the number in the income statement matches the number in the note that references it, which sounds trivial until you have watched three humans fail to keep it true across forty pages. It standardizes the formatting - the dollar signs, the commas, the spacing - which Antony says can eat up to 20% of an accountant's time all by itself. And it keeps an audit trail, linking each figure back to its source, so that speed does not come at the expense of the one thing accounting cannot compromise on, which is being able to prove where a number came from.

The design philosophy, as Inscope's investors describe it, is "AI-native": intelligence embedded directly in the workflow rather than bolted onto a spreadsheet. The subtle idea underneath is that a financial statement should behave more like software than like a document - a live object where a change upstream propagates automatically, instead of a static file that thirty-one versions of live in an inbox. Anyone who has played version-control roulette with a workbook named FS_Q3_FINAL_v7_KG_edits.xlsx understands the appeal immediately.

The money, and what it signals

In February 2026, Inscope announced a $14.5 million Series A led by Norwest Venture Partners, with Storm Ventures joining and existing backers Lightspeed Venture Partners and Better Tomorrow Ventures returning. That brought total funding to $18.8 million, including a $4.3 million seed round Lightspeed led back in 2023. The more interesting numbers are the operating ones: the company says it grew its customer base more than 5x and its annual recurring revenue more than 30x over the preceding twelve months.

You should always be a little skeptical of multiples off a small base - 30x is a very large number that can describe a very small starting point, and startups quote the flattering ratio. But the direction is real, and the customer list is the tell. It runs from CohnReznick, a top-15 national accounting firm, to fast-growing companies like Grubhub, Intercom, Guild, Hadrian, Synthesia and Oura. Accounting firms and their clients rarely agree on tooling; the firm wants control and the client wants convenience. When both show up on the same homepage, it suggests the product is solving a problem that exists on both sides of the engagement letter.

The bear case, stated fairly

The obvious risk is the one hanging over every "LLMs for a regulated, high-stakes workflow" company: hallucination is unacceptable when the output is a financial statement an auditor will sign and a lender will rely on. A chatbot that invents a citation is embarrassing; a reporting tool that invents a reconciling item is a restatement waiting to happen. Inscope's answer is the audit trail and the consistency checks - constrain the model, link everything to source, keep a human in the loop for judgment. Whether that is enough to earn durable trust is the thing the next few years will decide. The company is also entering a market with entrenched incumbents in disclosure management and a landscape of manual workflows that are annoying precisely because they are familiar and, in a grim way, trusted.

Still, the shape of the opportunity is legible. The work is repetitive, it recurs on a calendar, it is drowning in unstructured data, and it is staffed by expensive professionals who would rather be doing analysis than formatting. That is close to a textbook description of what vertical AI is supposed to eat. Inscope's wager is that the least-loved week on the finance calendar is also the most automatable - and that the people who lived through enough of those weeks are the right ones to build the escape hatch.

Selling to two customers at once

One underrated thing about Inscope's go-to-market is that it has to please two masters who do not always want the same thing. On one side are the accounting firms - CohnReznick, Armanino, Aprio, Connor Group - who prepare statements on behalf of clients and live or die by throughput and accuracy. On the other are the in-house finance teams at companies like Grubhub, Intercom, Guild, Hadrian, Synthesia and Oura, who want to close their own books faster and hand auditors something clean. A tool that speeds up the firm can, in theory, threaten the firm's billable hours; a tool that speeds up the client can, in theory, reduce what the client needs from the firm. Inscope's answer is that the busywork it removes was never the valuable part. The judgment - the materiality call, the disclosure wording, the sign-off - stays with the professional. What disappears is the copy-paste. If that framing holds, both customers win, which is the only version of this business that scales.

It also helps that the company is not asking anyone to abandon their system of record. Inscope reads from the ERP the customer already runs, most commonly NetSuite, and produces output in the formats auditors already expect. There is no rip-and-replace, no migration project, no year-long implementation - the things that kill enterprise finance software before it ever gets used. The wedge is narrow and specific: the stretch between "the numbers are final in the ledger" and "the statements are ready to sign." That stretch is where the spreadsheets multiply, and it is exactly where Inscope has planted itself.

The quiet ambition

The tagline - financial reporting so easy you might think it's fun - reads as a joke, and it is, but there is a real thesis underneath the wink. Antony's public framing of the company reaches beyond private annual reports toward the harder stuff: the 10-Ks and 10-Qs of public companies, the disclosures that get read by the SEC and short sellers alike. That is a much larger and much more dangerous market, where the cost of a wrong number is measured in restatements and lawsuits rather than embarrassment. Inscope is not there yet, and it is wise not to be. But the direction is clear enough. Start with the mid-market annual report, where the stakes are real but survivable, earn the trust that the audit trail is doing what it claims, and then move up the risk curve as the model and the process prove themselves. It is a patient plan for an impatient category, which is usually the plan that works.

aifintechsaasenterprisefinancial reportinggaapaudit-readynetsuitecontrollersgenerative ai