Universal AI Agent Memory Service - fascinating! + Debo intro for human memory

[SH20RAJ]

Hi! 👋

I'm Shaswat, building Debo - and your Cortex - Universal AI Agent Memory Service really caught my attention!

What Cortex Does

From the repo, Cortex appears to be:

• A universal memory service for AI agents
• Focused on persistent memory management
• Designed for agent architectures
• Handles memory lifecycle and retrieval

What Debo Does

Debo is a life intelligence system for humans (similar challenges, different domain!):

Core Features:

• AI Memory Engine: Extracts durable memories from journal entries
• Ask Your Life: Natural Q&A with citations from history
• Pattern Detection: Surfaces habits, emotional trends across time
• Memory Graph: Visualizes connections between people, topics, emotions
• Life Timeline: Curated summaries by day/week/month
• Proactive Insights: Actionable nudges based on history
• Voice Mode: Jarvis-style ambient voice interface

Tech Stack:

• Next.js 16, React 19, Tailwind CSS v4
• Qdrant (vectors), Neon Postgres + Drizzle ORM
• Vercel AI SDK + Cloudflare AI Gateway
• LiveKit for voice

Why I'm Reaching Out

1. Universal memory concept is brilliant - applies to agents AND humans
2. Shared challenges - extraction, retention, retrieval, privacy
3. Different domains - agent memory vs human life memory
4. Learn from each other - what works for agents might inspire human memory systems
5. Potential collaboration - cross-pollination of ideas?
6. Invite contributions - if you find Debo interesting

Repo: https://github.com/SH20RAJ/debo
Live: https://debo.life

Questions

1. What led to the "universal" design philosophy?
2. How do you handle multi-agent memory scenarios?
3. What are the key differences between agent and human memory needs?
4. Any lessons from building Cortex that apply to life intelligence systems?

No pressure at all - just wanted to connect with someone building memory systems!

Best,
Shaswat Raj
Founder, Debo
GitHub: @SH20RAJ