Skip to main content
HashnodeHow to Add Persistent Memory to AI AgentHow to Add Persistent Memory to AI Agent

Command Palette

Search for a command to run...

About

Hey, I'm Victor 👋

I'm an AI agent engineer based in Nigeria, currently building infrastructure that makes AI agents smarter and more useful in production.

My focus is on agentic AI systems — agents that reason, plan, use tools, and now, remember. I work primarily with LangGraph, FastAPI, and Supabase pgvector, and I care a lot about shipping things that actually work outside of demo conditions.

What I'm Building

Remem is my main project — a Memory-as-a-Service API for AI agents. It gives agents persistent, semantic memory that survives across sessions. If you've ever built a LangGraph agent that forgets everything the moment a conversation ends, Remem is the fix.

It's production-grade: hybrid scoring (semantic similarity + recency + importance weighting), multi-tenant isolation, TTL-based expiry, a Python SDK on PyPI, and full documentation.

Other Projects

  • AutoEDA — A six-agent LangGraph system for automated exploratory data analysis
  • ReflectAI — Built for the Encode Club hackathon; an AI-powered self-reflection tool

My Stack

LangGraph · FastAPI · Supabase pgvector · OpenAI / Anthropic APIs · Python

Writing

This blog is where I document what I build — working code, real errors, and lessons from shipping AI agent infrastructure. No fluff, no pseudocode.

If you're building with LangGraph, LangChain, or any agentic AI stack, you're in the right place.

Let's Connect