When Caitlin Leksana talks about the problem Fazeshift is trying to solve, she doesn't reach for Silicon Valley buzzwords. She talks about spreadsheets—endless, soul-crushing spreadsheets that finance teams use to track who owes what, who's paid, and who needs a reminder email. Again.
It's unglamorous work. But it's also, Leksana believes, exactly the kind of operational slog where AI can prove its worth beyond the hype.
On May 7, her San Francisco-based startup closed a $17 million Series A round led by F-Prime Capital, with backing from Gradient (Google's early-stage AI fund), Y Combinator, Wayfinder, Pioneer Fund, and a handful of other investors. The funding brings Fazeshift's total capital raised to $22 million, following a seed round announced in January 2025.
Founded in 2023, Fazeshift is building what it calls an "execution layer" for accounts receivable—essentially, autonomous AI agents that handle the tedious, repetitive tasks of chasing down payments, reconciling invoices, and nudging customers toward the pay button. The company says it's grown revenue 12-fold year-over-year and now counts dozens of enterprise customers, including eight unicorns and, recently, its first publicly traded company.
Named customers include Sigma Computing, Snyk, Meter, and Clipboard Health. Leksana says the platform has processed more than 9,000 customer communications in a single day. In one case, she says, a customer collected $7.4 million in outstanding receivables within weeks of switching on Fazeshift's system.
Whether that's impressive or table stakes depends on how much faith you place in AI's ability to navigate the messy, semi-structured reality of B2B payments. Fazeshift isn't the first to promise automation here—established players like HighRadius, Billtrust, Versapay, and Tesorio have been at it for years. But Leksana argues her platform goes further, deploying AI agents that don't just flag overdue invoices but actually execute on them: sending follow-ups, reconciling payments, even reviewing contracts for payment terms.
Eight Modules, One Headcount-Saving Pitch
Fazeshift offers eight modules spanning cash application, collections, AI-powered contract review, credit management, billing, payment disputes, payment portals, and customer onboarding. It integrates with the usual suspects—Salesforce, HubSpot, DocuSign, Stripe, QuickBooks, NetSuite, Gmail—stitching together workflows that typically require multiple logins and manual data entry.
The company's case studies lean heavily on quantified efficiency gains. PDQ, a payments company, reduced manual cash application work by 75% and avoided hiring additional finance staff despite acquiring other businesses. AdVon, a marketing firm, reportedly saved more than 80 hours per month and tripled accounts receivable capacity per full-time employee. Mecon, another customer, saw its overall AR balance drop 20% after adopting the platform.
These metrics matter in an environment where CFOs are under pressure to do more with less, and where "AI-powered" has become both a selling point and a red flag. Fazeshift holds SOC 2 Type II certification and says it operates under a zero-data retention agreement with OpenAI, a nod to the security concerns that shadow any product processing sensitive financial data.
From Nuclear Submarines to Finance Automation
Leksana didn't take the most obvious path to fintech entrepreneurship. She previously founded Carma Community and spent time at Boston Consulting Group before heading to Harvard Business School, where she met co-founder Timothy Galvin. Galvin's background is even less typical: he's an MIT-trained nuclear submarine officer who earned the US Navy's Admiral Andrew J. McKee Award in 2015—a distinction given to the top nuclear propulsion engineer in the fleet.
It's the kind of résumé detail that venture capitalists love, perhaps more than they should. But Galvin's engineering discipline and operational rigor do seem well-suited to the task of wrangling AI agents through the bureaucratic thicket of enterprise finance.
The team has grown from two co-founders to approximately 26 employees spread across San Francisco and New York. Fazeshift participated in Y Combinator and has since forged partnerships with AppWrap (a NetSuite implementation partner), Trimble Viewpoint, and The Hackett Group to expand its reach into mid-market and enterprise accounts.
The Broader Play
Leksana hinted in a recent interview that accounts receivable is just the opening move. She's aiming for what she calls "autonomous finance"—a broader CFO suite that could eventually handle everything from budgeting to close processes. It's ambitious, maybe too ambitious. But then again, finance teams are notoriously slow to adopt new technology, which means there's a lot of ground to cover if you can get a foot in the door.
F-Prime partner Rocio Wu participated in the round, though the firm didn't offer public commentary on the investment. Gradient's involvement is notable: the Google-affiliated fund closed a $220 million flagship seed vehicle in March, underscoring its continued appetite for AI infrastructure and application-layer companies.
Fazeshift recently refreshed its branding and has been making the rounds at industry conferences—the AICPA CFO Conference, Sage Future—pitching finance leaders on the promise of delegation without disaster.
The company didn't disclose its valuation, which is typical for Series A rounds where founders still have leverage and investors aren't eager to telegraph their pricing. What's clear is that Fazeshift is betting that the path to autonomous finance runs through the unglamorous, error-prone workflows that most people would rather not think about.
Getting paid on time. Reconciling invoices. Chasing down late payments. If AI can handle that, maybe it really is more than hype.
