Sciloop creates expert-level STEM reasoning data for frontier AI labs. The company produces advanced math and physics problems designed by International Physics and Math Olympiad medalists that current leading AI models cannot solve, then sells this training and evaluation data to AI laboratories. On its own benchmark, frontier models score in the low single digits on the hardest problems. Sciloop was founded in 2025, is based in San Francisco, and is part of Y Combinator's Fall 2025 batch.
Sciloop
ActiveExpert STEM reasoning data for frontier AI labs
Total raised
$500K
1 round
Stage
Seed
Jan 2025
Team
1-10
since 2025
Pricing
—
Founded
2025
San Francisco, United States
Agent-ready
—
Expert-level STEM reasoning data for frontier AI labs
Advanced math and physics problems designed by Olympiad medalists
Problems created by International Physics and Math Olympiad medalists
Training data for frontier model development
Evaluation data and benchmarks for model assessment
Problems calibrated to be unsolvable by current leading models
Own benchmark where frontier models score in the low single digits
Sold directly to AI laboratories
12/100
Early
MCP server
Public API
Webhooks
OAuth 2.0
SDKs
No public agent surfaces detected yet.
Jan 2025 Seed $500K ● Y Combinator
Capital network
$500K raised ·1 backer·10 network links
- Backers1
- Shared portfoliocompanies these backers also fund
- Extended networkfunds that co-invest alongside them
Databricks
The data + AI company
AI AgentsAI Infrastructure
Figure AI
General-purpose humanoid robots
AI InfrastructureAI Robotics
Upscale AI
Pure-play AI networking infrastructure
AI Developer ToolsAI Infrastructure
Dash0
AI-native observability platform built on OpenTelemetry
AI InfrastructureAI Data Engineering
Noma Security
End-to-end security for agentic AI
AI InfrastructureAI for Cyber Defense
Ineffable Intelligence
An AI research company building a superlearner to achieve superintelligence through reinforcement learning
Foundation ModelsAI Infrastructure
- What is Sciloop?
- Sciloop creates expert-level STEM reasoning data for frontier AI labs, producing advanced math and physics problems designed by Olympiad medalists that current leading models cannot solve.
- Who designs Sciloop's problems?
- The problems are designed by International Physics and Math Olympiad medalists to ensure they are rigorous and genuinely difficult.
- How hard is Sciloop's data?
- On its own benchmark, frontier models score in the low single digits on the hardest problems, reflecting the intended difficulty.
- Who buys Sciloop's data?
- Sciloop sells its training and evaluation data directly to AI laboratories building frontier models.
Discussion
Sign in to join the discussion.
Sign inExplore more around Sciloop
Contextual paths to related AI startups, deals and rankings.