Pujan Thapa

AI/Machine Learning Engineer & Data Scientist

I design and deploy intelligent systems, build scalable data pipelines, and develop robust Python backends. Passionate about transforming complex data into actionable insights and creating AI-driven solutions that are both powerful and efficient.

Download CV
A portrait of Pujan Thapa

Education

2023 - 2025

M.S. in Data Science

University of Naples Federico II, Italy

Specialized in Computer Vision and NLP. Thesis on "Computer Vision for Autonomous Driving".

2015 - 2019

B.S. in Computer Science and Information Technology

Tribhuvan University, Nepal

Work Experience

May 2024 - July 2024

ML Engineer, Omdena

Researched and developed solution during the timeframe of 8 weeks for Omdena Challenge from the data collection phase using Google Earth Engine, exploratory data analysis, model development using both supervised and unsupervised learning and model deployment in Streamlit Community Cloud.

Dec 2020 - Aug 2023

Software Engineer, Leapfrog Technology

Designed and deployed scalable Python-based ML, data engineering, and backend systems using AWS, Airflow, Docker, and CI/CD, with robust monitoring and automation. Optimized data pipelines, re-architected workflows, and improved test coverage with pytest, Selenium, and JMeter.

May 2018 - Aug 2018

Software Engineer, Parentiv Inc.

Researched and developed the Parentiv app’s backend architecture with Python (FastAPI), Google Cloud Firestore, and App Engine, integrating data from devices like Apple Watch and Google Home. Implemented NLP preprocessing with NLTK and spaCy, and ensured quality through unit and regression testing using pytest and JMeter.

My Projects

Project 1 Screenshot

CREA3 (In Progress)

CREA3 (Conflict Resolution with Equitative Algorithms) — An EU-funded project with the University of Naples Federico II and partners, focused on modernizing civil law procedures across Europe. Delivers faster, fairer, and more inclusive access to justice through electronic signatures, secure video conferencing, automated evidence logging, and support for users with limited digital literacy.

Python FastAPI PostgreSQL
Project 2 Screenshot

Plant Disease Detection

Developed a plant disease detection system using image analysis. Leveraged YOLO for object detection, alongside techniques like Non-negative Matrix Factorization (NMF) and Fuzzy Clustering for feature extraction and classification.

Python Roboflow Tensorflow HuggingFace
Project 3 Screenshot

Political Leaning Detection in News

Developed a web application to classify news articles as left-leaning, centrist, or right-leaning using fine-tuned BERT models. Enhanced model efficiency through quantization and Low-Rank Adaptation (LoRA). Deployed the solution using React JS, FastAPI and Docker, hosted on Azure Container Apps for real-time predictions, supporting media bias detection and analysis.

React Python (FastAPI) Docker Azure

Research & Publications

In Progress

Interests & Pursuits

Work Interests

Design Systems, AI/ML Architecture, Developer Tooling, Web Performance, and creating accessible digital experiences.

Research Interests

Human-Computer Interaction, 3D/4D Computer Vision, Information Visualization, and the societal impact of technology.

Personal Interests

Traveling Hiking/Trekking Bike Rides Gaming Cycling

Get In Touch

I'm always open to discussing new projects, creative ideas, or opportunities to be part of an ambitious vision. Feel free to reach out.

Say Hello