// Hey, I'm a — AI Engineer & Full Stack Developer
Ajay
Chowdary
I build production AI systems and full-stack applications — from LangGraph agentic pipelines and RAG platforms to React and Next.js products that ship and scale.
// Interactive
Interfaces with
intelligence
I build products where the experience and the engineering meet — agentic AI under the hood, immersive, responsive interfaces on top. This scene is live and interactive; drag it around.
// Skills
AI / Machine Learning
- Python
- PyTorch
- PyTorch Geometric
- GraphSAGE / GNNs
- LangChain
- LangGraph
- OpenAI API
- Claude API
- Gemini
- Hugging Face
- LoRA Fine-Tuning
- Prompt Engineering
- LangSmith
- Evals
Agentic & Orchestration
- LangGraph
- MCP
- Multi-Agent Systems
- Tool Use
- Function Calling
- State Machines
- Workflow Orchestration
- n8n
- CrewAI
Vector DBs & RAG
- FAISS
- pgvector
- Pinecone
- Weaviate
- Chroma
- RAG
- Embeddings
- Semantic Search
Frontend
- React
- Next.js
- TypeScript
- Tailwind CSS
- HTML
- CSS
Backend
- Node.js
- FastAPI
- PostgreSQL
- pgvector
- Prisma
- Supabase
- Auth.js
- Redis
- MongoDB
- Kafka
- Snowflake
Tools & Infra
- Docker
- GitHub Actions
- Vercel
- Git
- Stripe
- AWS
- GCP
- CI/CD
- Kubernetes
// Experience
Clemson University
M.S. Computer Science · Teaching Assistant
Aug 2024 — May 2026
Clemson, SC · GPA 3.76
- Completing graduate coursework in machine learning, algorithms, and AI systems while building production portfolio projects in LLMs, agentic workflows, and graph neural networks.
- Built a Heterogeneous GNN recommendation system achieving 57% Recall@10 vs 2% baseline — a 27× improvement — on 100,000 candidates using PyTorch Geometric and GraphSAGE.
Worthit Consultancy Services Pvt Ltd
Full Stack Software Engineer
May 2022 — May 2024
Remote
- Built and shipped production features across React and TypeScript frontends, Python and Node.js backends, and PostgreSQL data layers for 3 SaaS platforms — owning full product areas end-to-end with zero critical post-release incidents over two years.
- Designed developer-facing REST APIs with authentication boundaries, Pydantic-validated schemas, versioned contracts, and structured error handling — reducing API error rates by 25% and response times by 30%.
- Built LLM-powered automation workflows integrating OpenAI and Anthropic APIs into production operational processes, with evaluation frameworks tracking output quality before every deployment.
- Built GitHub Actions CI/CD pipelines with Docker and AWS deployments — reducing integration failures by 40% and manual release effort by 35%.
// What I Do
From model to product, end to end
- 01
AI Engineering
Design and ship agentic AI — RAG pipelines, LLM tooling, evaluation harnesses, and human-in-the-loop systems built to be trusted in production.
- 02
Full-Stack Product Development
End-to-end web applications, frontend to backend, optimized for speed, security, and scale with React, Next.js, and TypeScript.
- 03
ML & Data Systems
Recommendation engines, graph machine learning, and data pipelines that turn raw signals into measurable product value.
// Selected Work
A showcase of my latest projects
// Process
How I take an idea to production
Discover & Architect
Before a line of code, I dig into goals, data, and constraints — then design the system, the model, and the contracts between them.
Build & Iterate
Ship the interface and the intelligence in parallel. Every component — UI or model — is measured against real metrics, not vibes.
Launch & Improve
Deploy, instrument, and monitor. I keep iterating post-launch so the product gets sharper as it meets real users and real data.
// Writing