platoseedplatoseed
The foundational memory for Enterprise General Intelligence.
Enterprise General Intelligence starts with memory. Sentra captures the interactions where work actually happens, meetings, messages, emails & agent traces, and turns them into a shared organizational memory. That memory becomes the foundation for intelligent models that can understand and coordinate work across the entire company. As multiple time founders with backgrounds from MIT, IBM, and Caltech, we assembled a team of engineers and researchers from Google, Amazon, Meta, Microsoft, Apple, and Databricks to tackle this problem with us.
Sentra positions itself as a memory layer for teams and agents, capturing interactions, decisions, and drift, and modeling them as a single queryable graph. It integrates with many tools and enables querying in plain English with provenance attached to each answer.
Sentra builds a bi-temporal context graph by continuously syncing meetings, messages, docs, tickets, code, and CRM data from 200+ tools. It automatically extracts entities, relationships, and decisions without tagging or prompting. Facts have temporal validity (when true and when changed) and are linked to evidence. Users query the graph via API or MCP, with results grounded in the full context of the company and accompanied by citations to the source artifacts.
Who it’s for: Teams and enterprises seeking a unified, appliance-like memory layer for knowledge captured across tools and interactions; organizations needing persistent, provenance-backed context for agents and workflows.
Mentions product demo sign-ups, integrations with 200+ tools, and benchmark claims; indicates active development and enterprise focus
I was previously a professor at MIT and earlier at Caltech working on applied physics and AI, and also a researcher at Google. I created one of the early agentic LLM frameworks, Reflexion, and previously founded and exited two venture-backed companies. Currently exploring how organizational memory can enable Enterprise General Intelligence.
Previous founder, did AI engineering at IBM, Vapi, and ML research at MIT building offline deployments of ML systems for military use
Shared context for all of your team's agents.
Control what AI says about you
