In a bold leap for science and technology, Sciloop has emerged as one of the most promising AI-powered platforms redefining how research is conducted. The startup is automating the entire scientific process, from hypothesis generation to experiment execution and paper writing, all in record time.
Founded in 2025 by Bilal Asmatullah and Osman Siddique, both MIT affiliates and International Physics Olympiad medalists, Sciloop was born out of frustration with the slow and manual pace of traditional research. Their mission? To make scientific discovery faster, smarter, and far more accessible to researchers across the globe.
Built on Sakana AI’s Open-Source Framework
At the heart of Sciloop lies Sakana AI’s AI Scientist Framework, an open-source backbone that allows it to perform end-to-end research cycles autonomously. From generating hypotheses to running simulations, analyzing results, and formatting publication-ready papers, Sciloop can complete it all within minutes.
A demo video of the platform shows it executing a full infectious disease experiment using the SIR model and producing a polished academic paper, all for about 20 cents. To ensure data safety and high performance, the platform relies on managed cloud computing and isolated virtual environments.
Currently, Sciloop is in private onboarding with early adopters from MIT and other leading research institutions. Its main product, Sciloop Lab, can be accessed via sciloop.dev, where users can apply for early access or connect directly with the team.
Also Read: Oki AI Launches Knowledge Management Platform to Empower Growing Companies
Revolutionizing Machine Learning Research
Although Sciloop supports all scientific domains, its initial focus is machine learning (ML). Research teams using Sciloop can now automate tedious steps, such as data cleaning, code testing, and report generation, allowing them to focus solely on innovation and discovery.
Once researchers upload their goals and datasets, Sciloop’s AI runs parallel experiments, evaluates metrics, recommends refinements, and even drafts a complete scientific paper, including citations and references. What used to take months of work can now be done in just a few hours.
What sets Sciloop apart is its AI-powered peer review system. Another autonomous AI model critiques and evaluates the paper before approval, simulating real academic review standards while keeping the process efficient and low-cost.
From MIT to Y Combinator: The Journey Ahead
Sciloop’s acceptance into Y Combinator’s Fall 2025 batch marks a defining moment in its journey. Despite being run by just two founders, the startup has gained global attention for its “game-changing” impact on research productivity.
Inspired by a 2024 Stanford study showing that large language models generate more novel, though slightly less feasible, ideas than humans, Sciloop blends AI creativity with real-world validation, aiming for both innovation and accuracy.
The Future of AI-Driven Discovery
Looking forward, Bilal and Osman envision Sciloop not merely as a research tool but as a scientific co-pilot, one capable of contributing to breakthroughs in medicine, materials, and environmental science.
Though challenges remain in managing complex interdisciplinary research, Sciloop’s early success suggests a powerful new direction for how knowledge will be created, shared, and scaled.
In summary, Sciloop represents a turning point in scientific discovery, merging AI automation, human insight, and academic precision. With backing from Y Combinator and growing traction among MIT researchers, this platform is redefining what it means to “do research” in the age of artificial intelligence.
