Hello, I'm
Tinsae Befekadu
Designing autonomous multi-agent systems, enterprise copilots, and MCP integrations connected to scalable Node.js, Laravel, and Django backends.
Featured Work
A selection of production systems, cloud architectures, and scalable applications I have engineered.
Omishtu-Joy Tech Solutions
2026Omishtu-Joy Tech Solutions is an Addis Ababa-based software and AI engineering company that specializes in building enterprise applications, intelligent AI agent workflows, backend infrastructure, and specialized digital health platforms like the Healix+ Hospital Information Management System.
MeriAgent Hotel PMS and AI CEO
2026A hotel management platform with an AI executive co-pilot. Reservations, guest operations, pricing, and web content come together in one dashboard, with real-time insights to support revenue decisions.
DoxaRank AI SEO Platform
2026An Ethiopia-first AI-powered SEO platform engineered to track, analyze, and elevate search performance across Google Ethiopia and .et domains, combining technical audits, competitor insights, and agentic AI for automated search optimization.
About Me
Background & Engineering Approach
Building dependable AI agents on rock-solid backend infrastructure.
I am a backend developer turned AI agent engineer. Having built systems across Node.js, Laravel, and Django, I focus on practical engineering: deterministic tool execution, resilient API architectures, and reliable database layers.
Today, I design autonomous multi-agent pipelines with LangGraph and Model Context Protocol (MCP), backed by high-precision vector search in PostgreSQL (pgvector) and Redis. My focus is low latency, clean architecture, and software that simply works in production.
Core Technical Focus
LangGraph, State Graphs, Tool Calling, MCP Protocol, Agent Memory
Node.js, Laravel (PHP), Django (Python), REST APIs, Webhooks
Neon PostgreSQL (pgvector), Embeddings, Redis Caching, Hybrid Search
Celery Workers, Asynchronous Task Queues, Redis, Docker, System Design
Services I Offer
Comprehensive software engineering solutions to transform complex requirements into robust products.
Autonomous AI Agents & MCP Architecture
Designing autonomous multi-agent systems, stateful reasoning loops, and custom Model Context Protocol (MCP) servers that connect LLMs securely to enterprise databases and internal tools.
- Custom MCP Server Design & Endpoint Integration
- Multi-Agent Orchestration with LangGraph & State Graphs
- Deterministic Function & Tool Calling Workflows
- Structured Outputs & Token-Budget Optimization
Enterprise RAG & Semantic Intelligence
Engineering high-precision Retrieval-Augmented Generation (RAG) pipelines and vector search systems over complex enterprise documentation and domain knowledge.
- Sub-Second Semantic Search over Complex Data
- Vector Databases (Neon pgvector, Pinecone)
- Multi-Step Retrieval & Hybrid Search Pipelines
- Persistent Agent State & Memory Architecture
High-Performance Backend & API Systems
Architecting fault-tolerant backend services, high-throughput REST APIs, and asynchronous background pipelines to power high-concurrency applications.
- Scalable Backends in Node.js, Laravel (PHP) & Django
- Asynchronous Queues & Pipelines with Celery & Redis
- PostgreSQL / Neon Database Optimization & Migration
- Webhooks, Auth & Robust Production Guardrails
Work Experience
My professional engineering track record building autonomous multi-agent pipelines and scalable backends.
AI Agent & Backend Engineer Intern
Architecting autonomous multi-agent systems and task-execution pipelines that automate internal operations, client workflows, and document intelligence. Developing custom MCP server endpoints and high-throughput REST APIs in Node.js and Laravel with structured outputs and token-budget management.
AI Automation & Systems Engineer
Engineered autonomous SEO intelligence agents with LangGraph and MCP for tool discovery. Integrated real-time search analytics with Playwright automation, and architected asynchronous agent pipelines with Celery and Redis, backed by Neon PostgreSQL and pgvector for persistent agent memory.
Lets Create Together
Ready to bring your ideas to life? Let's discuss your project, architecture, or engineering opportunity.