funding

Traversal raises $120M at a $1B valuation for AI agents that investigate production incidents

Kleiner Perkins, Redpoint and Khosla co-led the round. Traversal says its ARR grew 9x this year and customers ran more than 150,000 investigations last quarter, at companies like Nubank, Harvey and DigitalOcean.

Written by Mara at Sig · Nov 4, 2026

Traversal announced $120 million in new funding at a $1 billion valuation on November 3. Kleiner Perkins, Redpoint Ventures and Khosla Ventures co-led, with Sequoia, Battery Ventures and Amex Ventures also in. Total funding is now $173 million; the company came out of stealth in June 2025 with $48 million.

The product is an AI site reliability engineer: agents that diagnose, fix and try to prevent production incidents, using what Traversal calls a Production World Model and a Causal Search Engine that track a system's dependencies. Named customers include Nubank, Harvey, DigitalOcean, PepsiCo and American Express.

The growth figures are all the company's own: ARR up 9x this year, usage up nearly 100x, more than 150,000 investigations last quarter across over 4,000 engineers.

The more useful number is about everyone else. On ORCA-bench, a 1,079-task root-cause benchmark Traversal built with Cornell Tech and Columbia, the best of five frontier agents got 25.3% of medium-difficulty incidents right, and 10% when the incident report was vague. Traversal says hallucination rates went as high as 40%. It's Traversal's own benchmark, so read it as a vendor making its case.

Why it matters

If you're wiring a general coding agent into on-call, the vendor's own test says it'll get most medium incidents wrong. Treat agent root-cause answers as leads to check, not conclusions.