Hone, a San Francisco startup building AI systems that take responsibility for business results, has raised a $60m seed round from Benchmark and Index Ventures. The company announced the round on 8 October 2026 together with the launch of its first product, which it calls Engines.
The round valued Hone at $285m, Bloomberg reported, using a figure supplied by the company. The report did not say whether the valuation was pre- or post-money, according to a round-up of coverage by Angel Investors Network.
Benchmark led the round with Index co-leading, FinSMEs reported, and Elad Gil, Hanabi, Definition and Diffusion also invested. Benchmark's Peter Fenton and Index's Shardul Shah have joined the board, according to the same round-up.
“What we are doing is building AI that owns outcomes.”Moritz Stephan, CEO, Hone
What Engines do
Hone describes an Engine as a persistent AI system assigned to an outcome, such as improving customer retention, rather than to individual tasks. Reports describe Engines being given roles such as procurement coordinator or outbound sales representative.
According to the company, an Engine first onboards itself by learning a customer's systems, people, history and constraints. It then works towards its goal, picking collaborators and staying within set guardrails, and keeps going after a given interaction ends. Hone says this is what separates Engines from chatbots and agents, which wait to be prompted.
The company says it judges success by the economic value an Engine produces rather than by demos, token use or the number of tasks completed. In its own research, Hone said, a replay-and-simulation approach cut errors from 33% to 5% across 125 internal test scenarios; the company ran those tests itself.
The team
Chief executive Moritz Stephan was previously chief of staff at Cognition, the AI coding company. Index wrote in its investment note that a friend had introduced its partners to Stephan as one of the most capable all-round people at Cognition, Angel Investors Network reported.
Hone says its staff come from Cognition, Decagon, Ramp, Mercor and Jane Street. A job listing earlier this autumn described a pre-launch team of fewer than 15 people, according to Founderland, as cited by Angel Investors Network. The company is hiring.
What the money is for
Hone said the funding is for building Engines. FinSMEs reported that the company will expand the platform and scale development of its autonomous systems.
Hone is exploring outcome-based pricing in which a customer sets a fixed budget, according to TBPN Digest. It has not named customers on its launch page, saying only that it is aiming at ambitious companies.
Stephan told Bloomberg the company's aim is to build AI that owns outcomes rather than one that completes tasks on request.
Hone did not disclose revenue, and beyond the job listing it has not said how large its team is. The final step in its own description of an Engine is delivery: the system reports results, keeps pursuing its goals and improves over time.
Company website: hone.com