Poetic lands $50M from Kleiner Perkins and OpenAI to run the enterprise work AI agents keep getting wrong — TFN

• Poetic raises $50M Collection A at a $500M valuation led by Kleiner Perkins, with OpenAI, Founders Fund, and First Harmonic additionally investing
• The San Francisco startup hit an eight-figure run charge and turned worthwhile in 2025 with simply 4 staff, automating fraud and compliance workflows for SoFi, AIG, and Chime. 
• Not like typical AI brokers, Poetic makes use of a proprietary programming language to show enterprise processes into deterministic, repeatable code. 

Most enterprise AI pilots die quietly. They appear spectacular in a demo, survive a managed check, after which stall the second they hit an actual compliance workflow, a fraud investigation queue, or a multi-hour insurance coverage assessment. 

Poetic, a San Francisco startup previously referred to as Forge, thinks it has cracked the issue, and a $50M Collection A suggests among the sharpest buyers within the enterprise agree. The spherical was led by Kleiner Perkins at a $500M valuation, with participation from OpenAI, Founders Fund, First Harmonic, and Genius Ventures. 

It’s not Kleiner Perkins’ first guess on Poetic: the agency additionally led the corporate’s seed spherical, and accomplice Leigh Marie Braswell personally backed the corporate earlier than it had a product to indicate. That form of conviction, doubling down at a ten-figure valuation, says one thing about what Kleiner Perkins has seen in manufacturing.

The chance is massive and largely unlocked

International monetary companies companies spend billions every year on guide operational processes, together with fraud investigations, transaction monitoring, compliance checks, insurance coverage evaluations, that current software program and AI instruments have didn’t reliably automate. The workflows are too advanced, too filled with unwritten guidelines, and too consequential handy off to a system that is perhaps proper 90% of the time.

Poetic was based in 2023 by Markie Wagner, a Thiel Fellow and Stanford pc science graduate who beforehand labored as an ML engineer at Google and Waymo earlier than founding Delphi Labs, an ML consultancy that deployed AI for among the world’s largest organisations. That background, seeing AI fail in manufacturing, again and again, throughout completely different industries, formed what she constructed subsequent.

The corporate’s core thought is a departure from the autonomous agent mannequin that almost all enterprise AI startups are chasing. 

Slightly than deploying a system that causes freely via a workflow, Poetic makes use of a proprietary programming language that lets operations groups describe advanced processes in pure language, then converts that logic into deterministic, repeatable execution. It learns like AI, however runs like code. The consequence, the corporate says, is automation that may deal with multi-hour processes working 1000’s of instances a day with near-zero tolerance for error.

The numbers in manufacturing are putting. At SoFi, Poetic reached 99%+ high quality executing fraud investigations end-to-end in 5 weeks. AIG and Chime are additionally named clients. 

The corporate says it has generated tens of tens of millions of {dollars} in financial savings for Fortune 500 purchasers and has transformed each pilot it has run right into a manufacturing deployment.

Probably the most aggressive corners of enterprise

Harvey⁠, which just lately raised $300M at a $5B valuation, is focusing on advanced authorized workflows. Abridge has raised $300M to automate medical documentation in healthcare. 

Most opponents, nonetheless, are constructing on high of general-purpose LLMs. which Poetic argues is the foundation reason behind the reliability drawback. Its language and deterministic execution mannequin use considerably fewer tokens, which the corporate claims makes it cheaper and extra controllable at scale.

“What Poetic has constructed is genuinely completely different — a platform that may execute the advanced, high-stakes processes that giant enterprises really run, with accuracy that exceeds what human groups can ship,” says Leigh Marie Braswell, Associate at Kleiner Perkins.

The larger image

The funding shall be used to increase Poetic’s forward-deployed engineering staff, enter new industries, and scale inside its current monetary companies buyer base. 

The corporate was worthwhile in 2025 on an eight-figure run charge with 4 staff. Headcount is anticipated to develop considerably because it strikes into new sectors. Complete funding up to now, together with the seed spherical, has not been disclosed.

The deeper query is whether or not Poetic’s strategy holds because it strikes past monetary companies into industries with completely different sorts of complexity — healthcare authorisations, authorities processing, provide chain selections. 

The corporate has chosen depth over breadth up to now. Whether or not that mannequin scales is the actual check forward.