Senior Software Engineer
Based in Kathmandu, Nepal
Building scalable systems and exploring deep learning, NLP, and music information retrieval. Currently working on Generative AI solutions at Verisk Analytics.
Based in Kathmandu, Nepal
Building scalable systems and exploring deep learning, NLP, and music information retrieval. Currently working on Generative AI solutions at Verisk Analytics.
I am a Senior Software Engineer at Verisk Analytics, specializing in building and scaling Generative AI and backend solutions. With a strong foundation in computer science and hands-on experience in production systems, I focus on bridging the gap between research and real-world applications.
My work spans from developing RAG pipelines and fine-tuning LLMs to creating scalable backend systems that serve AI models efficiently. I'm passionate about making AI accessible, especially for underrepresented languages and communities.
BSc Computer Science & Information Technology
Academia International College, Tribhuvan University
2019-2023 | First Division (74.24%)
Neural networks, CNNs, Transformers, adversarial robustness
Speech synthesis, text analytics, NLP for underrepresented languages (Newari, Rai)
Transformer-based models and LLMs for musical pattern analysis
RAG pipelines, LLM fine-tuning, production deployment
Verisk Analytics
Pathway Technologies
Led team of 7+ developers to launch a comprehensive travel marketplace platform.
Developed a secure and scalable application enabling efficient reporting, case management, and training for abuse prevention in sports.
Engineered advanced weather notification system and geo-mapping for agricultural app with 100k+ downloads.
Transformed years of Excel and handwritten government data, mapping potential Tube Wells and water resources using GIS.
Full-stack social media platform with real-time features using Django REST, Django Channels, WebSockets, and React. Complete separation of concerns with backend API and React frontend.
Real-time multiplayer game using Django Channels and WebSocket protocols for distributed state synchronization with sub-second latency. Deployed on Heroku with Redis.
Lightweight Object-Relational Mapping library for PostgreSQL with Nepali romanized terms. Published as PyPI package to make database programming more accessible for Nepali developers.
Blockchain waste management platform using Hyperledger Iroha with CNN-based waste classification using ResNet. Built during Ace Ignite 2023 hackathon.
Final Year Project | BSc CSIT 2023
Developed a full-fledged e-commerce platform with intelligent image search capabilities using PyTorch, Convolutional Neural Networks (CNN), and Django. The system enables users to search for products using images, combining computer vision with practical web application development.
App Development Competition Winner | Project Link
Won first place in a college-wide mobile app competition by building a cross-platform React Native application with Firebase, featuring a complete student information system with real-time data sync, authentication, notifications, and data management.
Python
JavaScript
TypeScript
PyTorch
CNNs & Transformers
LangChain & LangGraph
Generative AI
Django & FastAPI
Flask
VueJS & ReactJS
AWS (EC2, Lambda, ECS)
Docker
PostgreSQL & MongoDB
Redis
I'm always interested in collaborating on AI research projects, discussing new ideas in deep learning and NLP, or exploring opportunities in building production AI systems. Feel free to reach out!