Building high-performance Go backends, scalable Next.js frontends, and production-grade AI systems — from fraud detection engines to enterprise RAG platforms.
"I build systems that are secure by default, fast by design, and honest about what they do."
I specialize in the full delivery lifecycle — from schema design and Go API architecture to shipping polished Next.js interfaces. My work sits at the intersection of performance, security, and AI integration.
I'm completing a B.Tech in Cybersecurity at FUTO, which shapes how I approach every system I build: authentication surfaces, data exposure, injection vectors — security isn't a layer, it's a constraint I start with.
On the AI side, I've built production RAG platforms using semantic vector search with Llama 3/4 via Groq, and real-time fraud detection engines with rule-based scoring. I care about AI that's explainable and scoped — not magic boxes.
Built Go backend services and modern JavaScript web applications with cloud integration. Owned the full stack from API design to frontend implementation across multiple client projects.
Implemented secure authentication systems (OAuth, JWT) and shipped AI-powered features using third-party APIs, contributing to both internal tooling and client-facing products.
Designing and shipping personal and freelance projects spanning enterprise RAG platforms, fraud detection systems, real estate tooling, and developer CLI utilities. Each project treated as a production deployment — not a prototype.
Purposefully built. Each one solving a specific, real problem.
Enterprise RAG platform that turns static PDFs into a living, queryable knowledge base. Uses semantic vector search with intelligent text chunking to surface context-precise answers through a chat interface.
Built for teams that need private document intelligence without shipping data to external services. The indexing pipeline handles chunk overlap, embedding normalization, and confidence scoring before any query hits Llama.
Real-time fraud detection system analyzing financial transactions with rule-based risk scoring. Outputs Low / Med / High classifications with AI-generated natural language summaries for high-risk events.
CLI debugger that pipes build output to AI models for actionable fix suggestions. Auto-sanitizes 20+ sensitive patterns before any data leaves the machine, with smart token truncation for context limits.
Real estate platform with automated agent/landlord role detection, dynamic property listings, real-time Firestore updates, and click analytics for premium listings.
Open to fullstack contracts, AI integration projects, and long-term collaborations. I respond within 24 hours.