Software engineer specialising in backend systems, AI integration, and data engineering. Building production-ready applications from Kathmandu — open to remote roles and relocation.
I'm a Computer Engineering graduate from Kathmandu with a sharp focus on backend architecture, AI integration, and data-intensive systems. I love the intersection where clean engineering meets intelligent automation.
During my degree, I shipped a production ride-sharing backend on contract — building the full stack from WhatsApp chat-flow orchestration to Debian VPS deployment with Systemctl. Simultaneously, my final-year project achieved 85% directional accuracy in Nepal Stock Exchange prediction using Temporal Fusion Transformers.
I'm experienced across the modern backend stack: Node.js, Django, PostgreSQL, MongoDB, with growing depth in ML pipelines, LLM tooling (LangChain, Gemini API), and data engineering. I hold a Fuse Machines AI Microdegree and a DeepLearning.AI LangChain certification.
I'm looking for a role at a startup building something real with AI — where I can own backend systems, push ML integrations to production, and grow fast alongside sharp engineers.
AI-driven investment platform for the Nepal Stock Exchange. Built the full Django REST backend and PostgreSQL data layer, automated ETL pipelines processing 5+ years of OHLCV data across 200+ securities, and trained a Temporal Fusion Transformer for multi-horizon stock prediction. Integrated OpenAI Gym for exploratory RL-based portfolio optimisation experiments.
End-to-end LLM-powered document understanding system. OCR-based PDF ingestion, topic extraction via NLP, and automated flashcard and assessment generation using Google Gemini API. Secure Django REST backend with JWT authentication and PostgreSQL for longitudinal learning analytics.
Rigorous benchmarking of CNNs, Vision Transformers (ViT), and EfficientNet on a 50K+ plant disease image dataset. Applied transfer learning, data augmentation, and hyperparameter tuning. Evaluated across accuracy, precision, recall, F1-score, and inference latency with structured comparison reports.
Interactive web-based tool to visualise and compare sorting algorithms in real time. Dynamic controls for array size and animation speed — a clean demonstration of algorithmic thinking and vanilla JavaScript capability.
Open to backend engineering, AI/ML, and software engineering roles — at startups that are building seriously. Remote-first or US relocation. Visa sponsorship required.
Based in New Baneshwor, Kathmandu.
Open to remote work and willing to relocate .