Sciloop
Sciloop is a YC Fall 2025 (F25) company. Based in San Francisco, CA, USA. Team of 2.
About Sciloop
Summary
Sciloop builds an AI co-scientist that automates the experimentation and analysis loop for machine learning researchers. It replaces the manual work of designing, running, and interpreting ML experiments, letting researchers test ideas at a pace that current tools cannot match. The target user is the ML researcher who today loses weeks to workflow friction that sits between a hypothesis and a result.
Business model
Per-problem payments to experts ($225+ per problem) and dataset sales to enterprises.
Go-to-market
ML researchers and enterprises access Sciloop's platform directly; the site offers both a self-serve expert network and an enterprise dataset procurement path.
From the Y Combinator listing
AI Co-Scientist that automates ML experimentation and analysis.
Sciloop is building an AI Co-Scientist that automates experimentation and analysis for machine learning researchers, allowing them to test ideas far faster than today’s tools allow. We are Bilal and Osman, International Physics Olympiad medalists and the only Pakistanis admitted to MIT in our respective years, with several years of hands-on ML research experience inside MIT CSAIL. We have built large-scale training pipelines, worked with terabyte-scale datasets, and seen the research bottleneck firsthand while contributing to scientific ML projects. We left MIT to solve this problem directly and are now building a system that can eventually accelerate scientific discovery itself. https://sciloop.dev
Facts
- B2B / B2B -> Productivity
- San Francisco, CA, USA
- 2
- Active
- Early
Founders (2)
- Bilal Asmatullah
Score every batch against your own thesis
GPAgent's sourcing agent reads every company in a batch, scores each one against your fund's thesis in a private workspace, and lines up the founders worth meeting before Demo Day. You decide who to contact. Your scores, notes, and pipeline never appear on these pages.
How Accelerator intelligence works