Skip to Content
Introduction

Engram Documentation

Memory infrastructure for AI agents. Engram stores complete, uncompressed conversation transcripts and makes them instantly searchable.

Getting StartedConnect in 2 minutes
API ReferenceAll 6 MCP tools with parameters and examples
IntegrationsClaude Desktop, Cursor, Windsurf, custom clients

Learn

ConceptsHow conversations, chunking, and search work
ArchitectureSystem design, request lifecycle, and why this stack
Use CasesAgent memory, support history, knowledge bases, audit trails
ExamplesEnd-to-end tutorials and walkthroughs
PatternsBest practices for tagging, batching, and search
ComparisonHow Engram differs from Mem0, Zep, Supermemory

Reference

AuthenticationAPI key format and security
Data ModelFull schema reference with field descriptions
Limits and QuotasRequest limits, storage limits, platform limits
SecurityTenant isolation, encryption, privacy
GlossaryDefinitions of key terms

Operations

TroubleshootingCommon issues and fixes
FAQFrequently asked questions

Project

RoadmapWhat’s built, what’s next
ChangelogRelease history
Last updated on