- PDF ingestion: pdfplumber → LangChain text splitter → Sentence Transformer embeddings
- Semantic retrieval from Qdrant vector database
- FastAPI backend orchestrating indexing and query-time retrieval
- Streamlit chat UI connected to the API layer
- Fully containerized with Docker Compose
Bibishika Pokhrel
/
Computer Engineering graduate from Nepal, building at the frontier of language models and intelligent systems — RAG pipelines, agentic architectures, and infrastructure that makes LLMs actually reliable in production.
Biratnagar, Nepal Remote-first
<img class="photo-real">
AI systems
mastered
specialisation
worldwide
Building intelligent systems
from the ground up
I'm a Computer Engineering graduate from Purbanchal University, Nepal. My degree gave me the foundations — algorithms, probability, computer architecture, linear algebra — but what drives me is applying that knowledge to systems that actually work in the real world.
My primary focus is making LLMs genuinely reliable: grounding responses in real documents through RAG, designing careful chunking and embedding strategies, and building the API infrastructure that holds it all together.
Increasingly, I'm drawn to agentic AI — multi-agent systems with LangGraph, tool-calling patterns, and orchestration layers that let models do meaningful work autonomously, without constant hallucination.
Skills & Expertise
Featured Projects
Education & Formation
Bachelor in Computer Engineering
Purbanchal University School of Engineering — Biratnagar, Nepal
Four years developing strong foundations in algorithms, systems programming, mathematics, and AI. Graduated with a focus on intelligent systems and software engineering, with final-year work centred on ML applications.
Fields of Interest
Let's Talk
I'm actively looking for AI/ML engineering roles. If you're building something in the LLM, RAG, or agentic AI space — I'd genuinely love to hear about it. Full-time, contract, or just a conversation. All welcome.
Open to
new opportunities.
Based in Biratnagar, Nepal. Remote-first. Especially interested in roles working on LLM systems, RAG architectures, or agentic AI platforms.