Erdős AI Lab - Frontier AI Research
Erdős AI Lab is a San Francisco based research collective of young AI researchers advancing the frontier of artificial intelligence, from mechanistic interpretability, reasoning, and the mathematical foundations of learning to vision-language models, federated learning, world models and robotics, AI for science, and neuromorphic computing. Founded in 2025 by Dawar J Deka and backed by Entrepreneurs First. Contact: contact@erdoslab.org.
Pages
Team
- Dawar J Deka, Founder & Lead AI Researcher (IIT Bombay). Aerospace engineer turned AI researcher, working across AI and finance. Founder of Erdős.
- Nilesh Sarkar, Founding Principal AI Researcher (Moog Inc · Erdős AI Lab). Researcher and engineer building products and research across AI for science and foundational AI: vision, language models, genomics, and protein folding. Portfolio.
- Sunny Gupta, Federated & Decentralized AI (Ex Google DeepMind · Ex IBM · Ex Samsung R&D). Privacy-preserving federated learning, multi-agent systems, and edge AI.
- Dr. Amit Sethi, Principal Advisor (Professor, IIT Bombay). Founder of the MeDAL lab (Medical Deep Learning & AI) at IIT Bombay. Works on neural architectures, datasets, and domain adaptation. Faculty page.
- Dhruv Agarwal, AI Researcher (Ex Visa). Stanford ASES Fellow. Won $100K in Google cloud funding and beat Anthropic's open recruitment challenge. Works on multi-agent orchestration.
- Zainab Irfan, AIxBio · Neuromorphic (Erdős AI Lab). AI and biology, focused on neuromorphic computing and brain-inspired learning.
- Hitesh Meena, Computational Biologist (IIT Bombay). JEE Advanced AIR-99. Genomics, proteomics, and structural biology; led iGEM IITB's wetlab to Silver in Paris.
- Mohith H, AI Researcher (IIT Bombay · Deep Forest Sciences). Differentiable DFT and ML interatomic potentials; also ships vision and language model products.
- Riya Bisht, Neuromorphic Computing (Erdős AI Lab). Spiking neural networks and event-driven computation.
- Swapnil Kokare, Mathematics of AI (Erdős AI Lab). Mathematical foundations of AI.
- Netan Mangal, Foundation Models & RL (Erdős AI Lab). Foundation models and reinforcement learning.
- Chidvilas, Incoming Quantitative Researcher (Jump Trading). IIT Bombay Institute Rank 1. JEE Advanced AIR-1 '23. Simon Marais '24 winner.
- Mangalam Kartik, AI Researcher · Founder (Tensor & Maxwell Corp). One of the youngest reviewers at ICML 2026. Builds machine intelligence for robotics and autonomous systems.
- Tanishq Mandhane, Quant Finance · Deep ML (Jane Street). JEE Advanced AIR-29. Quantitative finance and deep machine learning.
- Matam Kushaal, Incoming AI and Systems Researcher (Samsung Korea). System architecture, deep learning, and edge AI; vision-language-action models.
- Harshith Pendela, RL Engineer (IIT Bombay). Multi-agent reinforcement learning, robotic learning systems, and world models.
- Samarth Sirsat, AI Researcher (Erdős AI Lab). Deep learning fundamentals to applied AI systems.
Products
Machine-readable summary: /llms.txt