I build tools that have to live with people. Right now that is Marshmallow, an open-source memory runtime for agents, and the sourcing and evaluation systems a venture fund runs on at Biome, where I am the only technical person on the team. Before that: lead engineer on an LLM guardrailing platform at HCLTech, with five engineers to train and review; product lead at a social network as it grew to 2.5 million users; and founder of a college social network that was acqui-hired.
The question underneath all of it: how should intelligent agents, and teams of agents, remember, reason, and remain legible to the people using them?











