hiloop is a San Francisco startup building infrastructure for automated research - the compute and observability layer that lets AI agents propose, run, and verify their own experiments. Teams hand hiloop a task, a model or agent, and a metric; hiloop runs snapshottable, forkable compute where every run is captured as a single queryable trace, then returns the best verified improvement across training, data, prompts, tools, and systems. Founded in 2026 by Karan Brar and Thomas Boser, it is part of Y Combinator's Summer 2026 batch.
Zuzanna Stamirowska is the co-founder and CEO of Pathway, a Palo Alto-based AI infrastructure company building real-time data pipelines and post-transformer AI architectures. With a PhD in Complexity Science from Paris I Panthéon-Sorbonne and a background in maritime trade forecasting recognized by the U.S. National Academy of Sciences, she leads a team that counts NATO, Formula 1, and La Poste among its customers. Pathway's $10M seed round in November 2024 brought total funding to $14.5M, and the company's October 2025 launch of the Dragon Hatchling (BDH) - a brain-inspired, continual-learning architecture - positions it at the frontier of post-transformer AI.

Cameron R. Wolfe, Ph.D. is a Senior Research Scientist at Netflix's Globalization team and the author of Deep (Learning) Focus, a twice-weekly Substack newsletter with 60,000+ subscribers that translates cutting-edge ML research into approachable long-form essays. A Rice University computer science PhD, he has built a reputation as one of the clearest explainers of large language models, RLHF, and AI agents in the field, bridging academia and industry with methodical depth and intellectual generosity.

Josh Tobin is a machine learning infrastructure pioneer who spent three years as a research scientist at OpenAI - contributing to the famous Rubik's cube robot hand - before earning his PhD from UC Berkeley under Pieter Abbeel. He co-founded Gantry, an ML monitoring and continual learning startup that raised $28.3M, and created Full Stack Deep Learning, the first course focused on production ML engineering. His domain randomization technique, which transfers neural networks trained in simulation to the real world, has been cited over 600 times and reshaped how robotics teams build perception systems. He runs a newsletter focused on ML infrastructure and ops.