Looking for a job

Dipu Thakur

@diputhakur

Message

I build practical AI, RAG, and backend systems for real-world document, data, and vision problems.

Nepal

Message

What I'm looking for

I'm looking to build reliable, practical AI systems that solve real-world problems, especially across generative AI, RAG, machine learning, computer vision, and backend development.

I've contributed to Digital Wakil at Aankhijhyal Technologies, building LangChain and RAG features for a production Nepali Legal AI system.

I integrated LLM services into a FastAPI backend and added Redis caching, reducing average response latency by about 40% on repeated queries. I also took the system from development to production on a Linux VPS using Docker, Ollama, Uvicorn, Nginx, systemd, SSL/HTTPS, monitoring, logging, and GitHub Actions CI/CD.

I build practical AI products including EmoSense for real-time facial emotion detection, LearnMate AI for PDF and image question answering, and an AI data analyst toolkit for exploratory analysis and chart generation. My work uses Python, TensorFlow, OpenCV, FastAPI, LangChain, LangGraph, FAISS, hybrid retrieval, OCR, and vector databases.

As a freelance web scraping and data extraction developer, I've delivered client projects extracting and cleaning more than 100,000 records from e-commerce, real-estate, and business-directory websites while maintaining 95%+ validated data accuracy.

Experience

Work history, roles, and key accomplishments

Full-Stack Developer Intern

Aankhijhyal Technologies

May 2026 - Aug 2026 (3 months)

Contributed to Digital Wakil, a production Nepali Legal AI system, developing LLM features with LangChain, RAG with hybrid retrieval, and prompt engineering. Integrated LLM services into the FastAPI backend and implemented Redis caching, cutting average response latency by ~40% on repeated queries.

Python Lang Chain RAG Pgvector PostgreSQL Git Github Docker NodeJS VPS Prompt Engineering fastAPI Redis Nginx Ollama Systemd SSL Let'S Encrypt CI CD GitHub Actions

Freelance Web Scraping & Data Extraction Developer

Self-Employed

Delivered end-to-end web scraping projects for clients, extracting and cleaning over 100,000 records from e-commerce, real-estate, and business-directory websites. Built scalable Scrapy spiders and BeautifulSoup + Requests scrapers, handling pagination, sessions, and rate limits with retry/backoff logic.

Python Scrapy requests BeautifulSoup JSON Git Github