Expression of Interest
Engram · San Francisco
CHECKED 3H AGO
Frontier labonsite
At a glance
Highlights
- Build compact memories that capture user knowledge and history.
- Open a new axis of scaling by studying context at training time.
- Partner with AI leaders like Microsoft, Notion, and Harvey.
Ideal candidate
Curious builder passionate about memory-centric AI and making models truly personal. Thrives in early-stage ambiguity and wants to shape a new scaling paradigm.
Lab profile
About Engram
Training models that study your work and form compact memories to anticipate questions.
What they build
- Continually learning models for workspace understanding
- Engrams — compact memories capturing user knowledge
Known for
- Continual learning
- Persistent memory formation
- Workspace personalization