DEV Community

Cover image for MEMORA: AI PERSONAL ASSISTANT
Victor Egbo
Victor Egbo

Posted on

MEMORA: AI PERSONAL ASSISTANT

Redis AI Challenge: Real-Time AI Innovators

What I Built

I built a Second Brain for Real-Time Automation, powered by Redis and n8n — a smart memory engine that:

  • Captures, stores, and recalls human-like context from multi-session conversations
  • Automates business workflows in real time
  • Uses Redis as the core memory, delay timer, and trigger controller

It works across platforms like WhatsApp, and it's designed for use cases such as:

  • Order and delivery automation
  • Agent memory recall
  • Time-based workflows (e.g. "start in 1 hour", converted into seconds using Redis TTL)
  • Context-aware task delegation

The system dynamically routes messages based on role (Admin, Customer, Agent), and it persistently remembers chat states using Redis JSON. It’s capable of acting like a CRM, automation layer, and AI memory unit — all in one.

Demo

🚀 [Live Workflow Screenshot] https://drive.google.com/file/d/1mXRqrzDCtEMs4yr2IO4Wst5pp825vOF6/view?usp=drive_link
🎥 [Demo Video] https://drive.google.com/file/d/1NKhmdgOHxA9NFlbc8Xe_8cxHCKxz5pRE/view?usp=sharing
📸 Screenshots:

  • Redis key structure
  • Message context stored and recalled
  • Real-time delivery timers firing workflows via TTL expiration

How I Used Redis 8

Redis was at the core of the brain. I used:

  • RedisJSON to store structured session history, roles, and memory per phone number.
  • Redis TTL + Key Expiry to schedule timed workflows ("start in 1 hour" is auto-converted to seconds and scheduled in Redis).
  • Redis Streams for logging and chaining AI agent interactions.
  • Pub/Sub to push live triggers to n8n for follow-ups, reminders, or escalations.

This allowed me to build a modular second brain that remembers, reacts, and acts in real time — like a digital operating system for conversations.

Tags

redischallenge #devchallenge #database #ai

Top comments (0)