Loading
Final-year BCT student at IOE Purwanchal Campus, Dharan / TU. Full-stack AI engineer shipping production apps end-to-end. Open to full-time roles starting 2026.
I'm a final-year BCT student at IOE Purwanchal Campus, Dharan / TU. I build AI-powered systems end-to-end — from designing backend APIs and task queues, to wiring model inference and RAG pipelines, to shipping polished frontends and mobile apps.
My stack is Django REST + Next.js + Flutter on the application layer, with Anthropic Claude, LangChain, and vector databases for the AI layer. I've shipped production agentic systems with tool-calling loops, conversation memory, and structured output extraction — not just API wrappers.
Currently based in Bhojpur, Nepal. I'm looking for full-time roles where I can work on hard problems at the intersection of software engineering and AI.
Designing and shipping agentic AI systems — tool-calling loops, conversation memory, RAG pipelines with vector databases, and structured outputs. Production experience with the Anthropic Claude API across 7+ endpoints.
Django REST backends with Celery task queues, Next.js frontends with server components, Flutter mobile apps with offline-first Drift persistence. I own the entire request lifecycle.
Three production apps with live users — a gamified learning platform, an AI finance tracker, and a distributed LLM inference engine. Published on the Play Store with daily-active users.
A gamified educational platform that turns spelling, vocabulary and general knowledge into short, rewarding daily rituals.
Django
DRF
FlutterAn AI-powered personal finance tracker that turns raw transactions into clear, actionable insight — works fully offline on mobile.
DjangoDRF
Next.jsFlutter
Anthropic Claude API+3A growing collection of structured patterns for building agentic AI systems on top of Django and the Anthropic Claude API.
DjangoAnthropic Claude API
LangChainAn academic minor project exploring how to serve large language models across multiple nodes — sharding model weights, fanning out prompts, and aggregating results at the edge. Built as a hands-on study of the trade-offs between throughput, latency, and hardware utilisation in distributed inference.
Python
PyTorch
Ray
FastAPI
gRPC+1A small, opinionated stack I reach for daily — and the long-tail I can drop into a project when it's the right tool for the job.
Python
Dart
JavaScript
TypeScriptDjangoDjango RESTNext.jsFlutterFastAPIAnthropicLangChain
OpenAIPyTorch
PostgreSQL
SQLite
Docker
Git
Firebase
Redis
CelerygRPCRayNotes on AI engineering, distributed systems, and the practical decisions behind the projects I build.
A practical guide to structuring tool-calling agents, conversation memory, and RAG-ready interfaces on top of Django and the Anthropic Claude API.
How I built Finia's mobile layer with local persistence, sync strategies, and a clean architecture that works without a network connection.
Lessons from an academic exploration of tensor-parallel model sharding, prompt routing, and throughput vs. latency trade-offs across multiple nodes.
Actively building AI engineering skills — RAG pipelines, LangChain agents, and vector databases — on top of production experience with the Anthropic Claude API.
Open to full-time roles, freelance, and collaborations. The fastest way to reach me is this form or LinkedIn.