Hi, I'm Netra — AI/ML Engineer from Nepal 🇳🇵
AI/ML Engineer specializing in Generative AI, Computer Vision, and scalable AI architectures. I build production-ready systems—from document intelligence and healthcare AI to multi-agent workflows—and share practical insights through research and technical writing.
Open to freelance projects, research collaborations, consulting, and AI product development.
I collaborate with startups, researchers, and organizations on challenging AI problems, applied research, and production-grade AI systems.
Career
Zakipoint Health · Lalitpur, Nepal
Built healthcare RAG systems with document parsing, architected multi-agent AI systems for autonomous actions and retrieval, and developed an LLM-powered PDF extraction pipeline.
AITC International · Bhaktapur, Nepal
Architecting AI/ML solutions for CRM services including ATS scoring, a multi-agent assistant, and AI-assisted automated workflow systems.
Dlytica Academy · Lalitpur, Nepal
Making Data Science and AI/ML concepts easy to understand through practical, accessible teaching.
Lincoln University College · Kathmandu, Nepal
Delivered hands-on AI/ML training to BIT students covering Deep Learning, NLP, and GenAI — from core theory to real-world projects.
Zakipoint Health · Lalitpur, Nepal
Developed disease prediction engines, fine-tuned language models for chatbot workflows, and built table detection systems for document processing.
Treeleaf Technology Pvt. Ltd · Lalitpur, Nepal
Built production identity verification with face detection, liveness checks, and OCR, plus a CNN-BiLSTM handwritten recognition engine.
Treeleaf Technology Pvt. Ltd · Lalitpur, Nepal
Developed SVM document classifiers, a YOLOv4 logo detector, and OCR-based information extraction from complex PDFs.
Skills
Technical Skills
Soft Skills
Projects
Document parsing engine and multi-agent AI for retrieving member-specific healthcare data and claims from complex unstructured PDFs and structured databases.
Multi-agent AI that matches candidates to job descriptions and auto-generates JDs for any role — built for HR teams to cut manual screening time.
Collected 450+ character images from 2,000+ students; trained a CNN-based classification model for word recognition across the full Devanagari character set.
End-to-end system with deep learning liveness detection and anti-spoofing before performing face recognition against a vector database to mark attendance.
Stacked-LSTM and Random Forest trained on 20 years of NEPSE data to predict future stock index values — covering data collection through model deployment.
Production gRPC service covering face detection, liveness checks, OCR, NER-based information extraction, and document verification for onboarding workflows.
Blog
A hands-on walkthrough of Model Context Protocol — building a personal AI research assistant that connects tools, memory, and LLMs.
Read on Medium →Everything software engineers need to understand about Model Context Protocol — architecture, use cases, and practical implementation.
Read on Medium →A comprehensive guide to the most effective techniques for reducing hallucinations in large language models across real-world deployments.
Read on Medium →Hire Me
I help startups, researchers, and organizations design and ship production-grade AI systems — from initial architecture to deployed product.
Have a challenging AI problem? I'm open to freelance engagements, research collaborations, and consulting.
Email me View resume