Give Your Slack a Memory: A Cognee and Qdrant Hack Night
The Problem
Most agent memory today is a Markdown file the agent writes to itself. That works for one person on one laptop. It breaks as soon as the memory is shared across a team, longer than a context window, scoped by who is allowed to see what, and full of decisions that reference other decisions. A Slack workspace is all four at once.
Why Join
Cognee and Qdrant are teaming up for an evening of building a memory layer that survives those conditions. Cognee reads your Slack history and builds a typed knowledge graph from it: entities, relationships, and provenance back to the original message. Qdrant carries the retrieval load. Every agent turn hits the vector store, so two things matter. All of it runs on the hardware that you control, which matters when the corpus is your company's Slack.
You'll build the whole loop: ingest, structure, retrieve, then compile the patterns you find into agentic SKILLs your agent can reuse. Bring a laptop. You'll leave with a pipeline you can point at Jira, Notion, or your own codebase.
What You'll Build
Working in small teams (or solo, if you prefer), you'll wire up Cognee's memory engine with Qdrant's vector search to give a Slack workspace real recall. By the end of the night you will:
Ingest messages and threads from a Slack export or live workspace
Structure that history into queryable memory
Answer questions a plain keyword search never could
Show us what you got
You will have 2 minutes to present a demo in which you:
Present your idea - Explain what memory problem you solved and how you used Cognee and Qdrant to do it
Run a live demo - Show your build in action, solving something a plain search or a stateless bot couldn't
🏆 1,200€ Cash Pool
Main Prizes:
🥇 Prize: 600€ cash
🥈 Prize: 400€ cash
🥉 Prize: 200€ cash
Plus amazing swag from both Qdrant and Cognee for all the participants.
Schedule
17:00 PM - Doors open + Networking
17:30 PM - Opening remarks + Cognee & Qdrant intro
18:00 PM - Hacking begins
21:00 PM - Project Submission Deadline – finalist selection
21:15 PM - Finalist presentations & judging
21:45 PM - Awards ceremony
22:00 PM - Event wrap-up & doors close
Bring a laptop. We'll have API keys, sample datasets, and people from both teams on hand to help you get unstuck.
Limited spots available. Can't make it? Please cancel your RSVP so we can offer your place to someone else.
Photos will be taken throughout the event and may be used by Cognee afterward.
About Qdrant:
Qdrant is an open-source vector search engine built to handle high-dimensional data at scale. It powers the retrieval layer behind some of the most demanding AI applications in production today, from RAG pipelines to recommendation systems to AI agents, giving developers fast, accurate similarity search wherever it's needed.
About Cognee:
Cognee is an open-source AI memory engine that transforms unstructured data into persistent, queryable memory. It enables AI agents to store, retrieve, and reason over information across time and infrastructure, making it possible to build reliable, adaptive, and stateful AI systems that operate seamlessly across cloud and local environments.