Bibit Kunwar Chhetri
Skills
Languages: Python, Bash, SQL, Lua, LISP, JavaScript, Java, LaTeX, C/C++
Technologies & Tools: GNU / Linux, Docker, Kubernetes, Git, Terraform, Ansible, Prometheus, Grafana, Loki, LangChain, LangGraph, WordPress, IoT, Nvim, Emacs
Work Experience
Assistant Lecturer
- Deliver lectures on AI Application Engineering and Cloud Computing, focusing on practical and industry-relevant concepts.
- Design and guide hands-on projects involving large language models, APIs, and cloud deployment workflows.
- Mentor students on problem-solving, system design, and best practices in modern software and cloud-based architectures.
Software Engineer / System Administrator (Mid-Level)
- Worked as a System Administrator, System Designer, and Backend Engineer, handling end-to-end system architecture and infrastructure management.
- Designed scalable system architecture and backend services to support high-traffic web applications.
- Managed servers, deployments, and performance optimization to ensure system reliability and uptime.
- Developed and maintained backend APIs and services for the product nepalbahas.net (on-going).
- System Design, Backend Development, DevOps, Server Management
Software Development Engineer (Intern)
- Worked on backend development using Java and Spring Boot, contributing to RESTful APIs used in fintech applications.
- Built data scraping scripts using Python to collect and process financial data from external web sources, enabling automated data acquisition for internal analysis and reporting.
- Assisted in API development, validation, and debugging, ensuring reliable communication between backend services and client applications.
- Spring Boot, Selenium, Beautiful Soup, JPA/Hibernate, Liquibase, Maven
AI Application Development
- Developed an AI-powered automated risk monitoring system that fetches daily financial news, performs LLM-based sentiment analysis, and assigns contextual scores to identify potential risks.
- Built a semantic search and query interface using RAG and Pinecone Vector DB, enabling interactive exploration of entity-specific news through a Streamlit dashboard.
- Implemented automation pipelines for daily data ingestion, processing, sentiment scoring, and email reporting, reducing manual monitoring efforts and providing timely insights.
- RAG, Vectorless RAG Agentic Architecture Design, SerpAPI, Pinecone VectorDB, Streamlit
Data Extraction and Preprocessing
- Scraped US FOMC press release transcripts by building a subtitle extraction automation script and created a cleaned research dataset through preprocessing.
- Additional freelancing experiences: Economics Simulation Model in Python, SQL Agent for Simulated Banking Database, eCommerce WebApp for a clothing/restaurant business on WordPress.
Education
B.E. in Software Engineering CGPA: 3.34/4
Relevant Coursework: Operating Systems, Object Oriented Programming, Databases, Discrete Maths, Data Structures and Algorithms, Computer Networks, Machine Learning, Data Mining, Advance Data Structures and Algorithms, Information Retrieval, Image Processing, Organizational Management, Engineering Economics
Project Work
TriNetra: Real-Time Violence Detection System (2025): Developed a mobile-integrated CCTV monitoring system to automatically detect physical violence in real-time video streams using a hybrid deep learning model (MobileNetV2 + Bi-LSTM). Built a microservice-based architecture with FastAPI, Redis queues, and FFmpeg for video processing and alert notifications.
Python, FastAPI, TensorFlow, React Native, PostgreSQL, Redis, Docker, RTSP, FFmpeg
GENIE – Bilingual Voice Recognition Tool (2024): Built a bilingual Automatic Speech Recognition (ASR) system for Nepali–English mixed speech using a Transformer encoder–decoder architecture. Trained the model on the Crowd-Sourced Nepali Speech Corpus and Svarah datasets and deployed it as a web application.
Python, FastAPI, JavaScript, MySQL, Tailwind CSS
PicAtU: Face Recognition Photo Retrieval System (2023): Developed a web application that enables users to find and download event photos using facial recognition. Implemented face detection and encoding with the Dlib C++ library and designed a microservice-based system with React and multi-language backend services.
React, Node.js, Flask, MySQL, C++, Dlib, CUDA
Awards and Certificates
- Mentor - National Innovation Center, Pokhara: Helping students and working professionals to get better at Linux fundamentals and get them ready for EHR project.
- Supervised Machine Learning: Regression and Classification - Coursera (Stanford Online), DeepLearning.AI
- Annual Nepal AI School (ANAIS 2023) - Organized by NAAMII – NepAl Applied Mathematics and Informatics Institute
- Volunteer/Organizer of WordCamp Nepal, 2024 - Organized international event in Pokhara
- Runner-Up – GCES IT Expo: Organized by GCES IT Club