Open to opportunities

Full-Stack Engineer specializing in real-time systems and AI integrations Building production-grade applications with strong backend architecture and low-latency systems

I'm a full-stack engineer with 3+ years of experience building and shipping production applications, with a focus on backend systems, real-time infrastructure and AI integrations.

3+ years experience
Owned end-to-end delivery

About

Building intelligent systems with production discipline and systems thinking.

I'm a full-stack engineer with 3+ years of experience building and shipping production applications, with a focus on backend systems, real-time infrastructure and AI integrations.

I work primarily across the MERN stack, designing systems that handle live data, scale under load and integrate LLM capabilities where they add real value.

Focus Areas

  • Real-time systems (WebSockets, Redis pub/sub, low-latency sync)
  • Backend architecture (caching, rate limiting, fault tolerance)
  • Practical AI integration (RAG pipelines, prompt workflows, API orchestration)

Core Stack

JavaScriptTypeScriptReactNode.jsExpressMongoDBGitPostgreSQL

Real-Time & Backend

WebSockets (connection management, sync strategies)Redis (caching, pub/sub)Rate limiting, queue-based processingAPI design (REST + real-time patterns)

AI Integration

OpenAI, Claude, Groq APIsRAG pipelines (basic retrieval flows, embedding-based search)Prompt design and response validationLatency + cost-aware API usage

Infrastructure & Production

DockerAWS (deployment, scaling basics)Observability (logs, metrics)CI/CD pipelines

Soft Skills

CommunicationLeadershipSystems ThinkingEnd-to-End OwnershipBias for ActionInfluence without AuthorityAmbiguity Management

Certifications

Building with the Claude API2026

Real-Time Systems Thinking

Beyond REST APIs—I build WebSocket-driven collaboration systems with conflict resolution, presence awareness, and sub-100ms latency. Redis pub/sub, message queues, connection pooling.

System Architecture

Designed WebSocket architecture supporting 1K+ concurrent users with Redis pub/sub. Implemented caching layers reducing DB load by 60%. Built fault-tolerant services with retry + fallback mechanisms.

Production Discipline

Code that ships, scales, and survives. TypeScript strict mode, comprehensive error handling, database indexing strategies, caching layers, CI/CD pipelines, rate limiting, and observability with logs/metrics/tracing.

Flagship Projects

Deep dives into systems I've built. Click to expand for architecture details, challenges solved, and measurable impact.

AI-Powered Learning Platform

Adaptive learning paths driven by LLM integration

ReactNode.jsMongoDBSupabaseRedis+2 more
Thousands
Peak Users
~180ms
Avg Response Time
Improved
Completion Rate

Real-Time Collaboration System

Multiplayer workspace with conflict resolution

ReactNode.jsWebSocketRedisMongoDB
~1K
Concurrent Connections
<50ms
Sync Latency
99.9%
Uptime

Additional projects available on request

Experience

A track record of shipping production-grade software and growing technical capabilities.

2025

Full-Stack Engineer

Cytonn

  • Built and maintained fintech platforms serving thousands of users
  • Developed backend services with Redis caching, reducing API latency significantly
  • Implemented real-time features for live market data using WebSockets
  • Improved system reliability through structured error handling and monitoring
2024-2025

Data Annotation Specialist

Contracts

  • Worked on large-scale datasets for ML systems (50K+ items)
  • Identified inconsistencies in data pipelines and improved processing efficiency
  • Collaborated with ML teams, gaining exposure to model behavior and evaluation
2022 — 2023

Freelance Junior Developer

Freelance

  • Developed MERN stack applications from scratch for 5+ clients
  • Integrated OpenAI API for automated content generation features
  • Implemented RESTful APIs and real-time features using WebSocket

Let's Build Something

Looking for a FullStack engineer who can architect AI-powered systems and ship production-grade code? Let's talk.

Response time: usually within 24 hours