There's a problem lurking in the enterprise software stack—one that analysts peg at more than $89 billion when you count implementation, upgrades, and support for SAP's S/4HANA transition—and it comes with a deadline that can't be extended indefinitely.
Nova Intelligence, a two-year-old startup building AI agents to automate SAP system migrations, announced May 5, 2026 that it had closed $31.5 million in Series A financing. Chemistry led the round, joined by Accel, Conviction, and—perhaps most tellingly—SAP.iO, the venture arm of the very company whose legacy infrastructure Nova is racing to modernize. The funding brings the company's total raised to north of $40 million, following an $8.3 million seed round reported by Forbes in January.
The timing matters. SAP's aging ECC platform, which as of recent years processes 77% of the world's transaction revenue, faces mainstream maintenance cutoff at the end of 2027. Extended support runs through 2030, but that's a narrow window for thousands of enterprises to rework custom code and migrate to S/4HANA. Which is to say: this isn't optional. It's a forced march.
When SAP Itself Decides to Back You
Kristina Shen at Chemistry spearheaded the Series A, while Ivan Zhou from Accel—who led Nova's seed—doubled down. But the real signal came earlier, in June 2025, when SAP.iO publicly confirmed its stake. Having the platform owner invest in a company automating migrations off its own legacy systems suggests validation, though perhaps also pragmatism. SAP knows the migration bottleneck could become its own liability.
Conviction, the AI-focused firm Sarah Guo founded, rounded out the investor group—a logical addition given Nova's bet on agentic architecture rather than traditional consulting labor.
The Team That Actually Knows SAP's Guts
Nova's founding trio brings credentials that read less like a typical startup résumé and more like a targeted strike team.
CEO Emma Qian spent time as a research engineer at Google DeepMind and Meta AI. CTO Sam Yang previously built Bloom, a consumer fintech app that pulled in more than two million users before he pivoted to enterprise infrastructure. But the marquee hire is Chief Scientist Alexander Zeier, who co-invented SAP HANA itself and later served as Global CTO of Accenture's SAP Business Group. It's the kind of insider credibility that opens doors—and makes early customer pilots considerably easier to secure.
Early Numbers That Beg Scrutiny (in a Good Way)
Nova is making bold productivity claims, and so far, customers seem willing to attach their names to them.
Festo, a manufacturing automation company, claimed five-times productivity gains across development tasks in a May 2026 case study published by Nova. KION Group, an industrial logistics firm, reported 400% performance improvements in the company's case materials. E-3 Magazine, an independent SAP trade publication, dug into KION's deployment and reported application runtimes dropping from roughly 25 minutes to two or three minutes, with code analysis cycles shrinking from one to two weeks down to two to four days.
A Q2 2025 pilot with Kyndryl showed a 75% reduction in manual implementation effort and 50% cost savings while modernizing around 33,000 lines of ABAP code. That's meaningful, if it scales.
The company's operating footprint remains compact—Fortune pegged the team at around 20 people in May 2026, though LinkedIn suggests the headcount might range between 11 and 50. Nova has formalized partnerships with Kyndryl (announced in August 2025) and Mindset Consulting (May 2025), presumably to distribute its platform through existing consulting channels.
The Agentic Bet
What Nova is building isn't just migration tooling. The company's platform handles functional design, fit-to-standard analysis, custom code modernization from legacy ABAP to newer frameworks like RAP and Fiori, automated testing, and production issue resolution. In other words, the entire modernization lifecycle—automated by AI agents rather than armies of consultants billing hourly.
"We're betting agentic AI can compress this work dramatically," Qian told Fortune.
The business model reflects that thesis. Instead of pricing per object modernized—the traditional consulting approach—Nova charges annual platform fees plus per-seat licenses. It's a structure that scales with agent capacity, not human labor. Whether enterprises will accept that shift in economics remains an open question, but the compressed timeline may not give them much choice.
Nova plans to deploy the fresh capital on R&D and expanding its go-to-market motion, targeting what the company calls "flagship enterprise customers." Given SAP's installed base and the looming deadline, the addressable market is enormous. The question isn't demand. It's whether Nova's agents can deliver at the pace and scale the calendar requires.
The clock, after all, is already ticking.
