Jonathan Chavez said the team stopped working on ZeroEval and sunsetted the product for customers, including DoorDash, while shifting focus to LLM Stats.
SourcePublic company, workplace, funding, and market signals
Updated Jul 29, 2026
YC-backed AI infrastructure company building an independent layer for evaluating and improving AI agents and models with human feedback; its first public product is LLM Stats.
Primary product
LLM Stats
Founded
2025
Headquarters
San Francisco, California, United States
Team size
2-10 employees
Industry
Software Development
Sub-industry
AI agent evaluation and optimization
Business model
0 jobs at ZeroEval
Check back later for new openings
Stage
Pre-seed
Total raised
$500K
Latest round
Pre-seed · Oct 2025
Latest amount
$500K
Oct 2025 · Y Combinator
Investors
Founder-led, highly technical, and customer-embedded; public messaging emphasizes fast iteration, reliability for production AI, and using human feedback to improve systems.
Differentiators
Technology
Customers
Competitors
Estimated revenue
≈$100K–$500K ARR (GetLatka estimate: $220K)
Jonathan Chavez said the team stopped working on ZeroEval and sunsetted the product for customers, including DoorDash, while shifting focus to LLM Stats.
SourceLinkedIn · May 2026
Jonathan Chavez said the team shut down the ZeroEval product for customers and moved focus to LLM Stats.
YesPress · Apr 2026
Profile article describing ZeroEval as a New York-based YC S25 startup, with product, customer, and team details.
Fondo · Feb 2026
Secondary launch write-up summarizing the product, use cases, and why teams would use it for agent evaluation and optimization.
LinkedIn / Y Combinator · Aug 2025
YC promoted ZeroEval's launch and reiterated the product's agent-evaluation and prompt-optimization pitch.
Y Combinator
Official YC launch describing ZeroEval as a tool for evaluating and optimizing AI agents with human feedback, including calibrated LLM judges and Autotune.