Hi, I'm Ujwal Dahal
A passionate AI/ML Engineer building intelligent systems, RAG pipelines, and LLM-powered applications with Python, LangChain, and cutting-edge AI tools.
Ujwal Dahal
AI/ML Engineer
About Me
Hi! I'm Ujwal Dahal, a passionate AI/ML Engineer with a strong drive for building intelligent systems that solve real-world problems. I specialize in machine learning, deep learning, and LLM-powered applications.
I have hands-on experience developing RAG (Retrieval-Augmented Generation) pipelines, LangChain applications, and NLP solutions. My background in web development gives me a unique edge — I can build full-stack ML applications from model training to production deployment.
My goal is to leverage data-driven approaches and cutting-edge AI tools to create impactful solutions. I'm constantly exploring vector databases, transformer models, and the latest advances in the LLM ecosystem.
My Skills
Python
TensorFlow / PyTorch
Pandas / NumPy
LangChain
Scikit-Learn
SQL
NLP
Streamlit
Git
Docker
My Projects
IntelliDoc
Multi-agent document intelligence platform — summarization, interactive mind maps, and AI debate analysis using LangGraph orchestration.
Nepali Sentiment Analysis
Fine-tuned XLM-RoBERTa on 65K Nepali-language examples achieving 88.5% accuracy vs 50.5% zero-shot baseline (+38 percentage points). Built with a manual PyTorch training loop — no Trainer API.
ConversationalRAG
Conversational document Q&A backend with session memory and structured intent extraction, built without LangChain's retrieval abstractions.