Hi Community,
We’ve just launched Memori — a human-like memory solution that works seamlessly with CrewAI to give your agents context awareness and continuity.
With Memori + CrewAI, AI Agent can:
Remember user preferences across sessions
Inject only the most relevant context automatically
Build more personalized, adaptive agents
Why Memori for CrewAI? Why is Memori Different?
-
Dual-mode architecture: Conscious Mode (short-term working memory) + Auto Mode (intelligent database search)
-
SQL-first design - works with SQLite, PostgreSQL, MySQL (no vector DB required)
-
Backed by a strong database infrastructure that comes with multiple database branching, database versioning, query optimization, point of recovery, and more…
-
Namespace separation for multi-tenant setups
-
One-line setup:
from memori import Memori
# Enables conscious mode
memori = Memori(conscious_ingest=True)
# Auto-records all conversations
memori.enable()
Example: I built travel planner agents in CrewAI that remember destinations, budgets, and past trips — delivering truly personalized recommendations without repeating prompts.
Get Started:
memori/demos/travel_planner at main · GibsonAI/memori · GitHub
Try it, give us a 
, and share your feedback — we’re improving Memori based on the community’s needs! Thanks so much!