Full Stack Products
Interfaces, APIs, and the workflow behind them, taken through to a deployed system.
Next.js · Node.js · APIs
Available for Full Stack and AI roles
Full Stack Engineer | AI/LLM Engineer
I build modern web applications and intelligent AI-powered systems, combining full-stack engineering with practical AI solutions.
From modern web applications and enterprise backend systems to AI agents, RAG pipelines and document intelligence, I engineer complete solutions that solve real-world business problems.
Surat, India · Open to Remote India, Delhi NCR (Delhi, Noida, Gurugram)
Synthetic output{
"doc_type": "vehicle_registration",
"language": ["ar", "en"],
"confidence": 0.97,
"valid": true
}3+ Years Experience
Full Stack Development
Applied AI Engineering
India / Open to Opportunities
02 / Work
03 / Practice
Interfaces, APIs, and the workflow behind them, taken through to a deployed system.
Next.js · Node.js · APIs
Tool calling, grounded answers, and a second model when the first one fails.
Tool calling · RAG · Handoff
Detect a card, read Arabic and English, and return fields a downstream system can validate.
YOLO · OCR · Validation
Claims, policies, roles, and the jobs that keep an insurance operation moving.
CRM · Claims · RBAC
04 / Stack
React.js · Next.js · TypeScript · JavaScript · Redux · Tailwind CSS · HTML · CSS
Node.js · Express.js · Python · FastAPI · REST API design · Authentication · Role-based access control
MongoDB · Pinecone · FAISS · Embeddings · Vector search
OpenAI API · Gemini · Grok · Prompt engineering · Structured outputs · Tool calling · Multi-LLM fallback · RAG · AI agents · LangChain · LangGraph · CrewAI
OCR · YOLO · OpenCV · Google Vision · PaddleOCR · EasyOCR · Document detection · Data extraction · Validation
Admin dashboards · CRM · Workflow automation · Database design · Third-party integrations
AWS · Git · CI/CD
05 / Approach
01
Fallbacks, retries, and validation come before a new model or a clever abstraction.
02
Models are for the steps that need judgment over language or documents. The rest stays ordinary software.
03
Insurance and customer decisions that carry real consequences get a handoff, not a silent guess.
04
A system another engineer can operate is worth more than a pipeline only its author understands.
LLM Engineering · 19 Nov 2026 · 6 min
How an application asks a model for a structured tool call, checks the arguments, runs the tool itself, and sends the result back to the model.
LLM Engineering · 19 Nov 2026 · 6 min
A practical way to call a second model when the first one times out or returns an unusable response, without repeating the same side effect.
RAG · 19 Nov 2026 · 6 min
How to ingest business documents, retrieve the right passages, and keep the written answer tied to those passages instead of model memory alone.
Sep 2023 — Present · Surat, India
End-to-end ownership of production insurance systems: a WhatsApp contact center, document OCR, policy microservices, and a claims CRM.