|
I'm studying Data Science at the University of Milano-Bicocca (UniMiB) in Milan, and I'm looking for data science and machine learning roles. I co-authored an IEEE paper on thermal-infrared human detection, built Power BI dashboards tracking 35,000+ service calls a month as a data analyst, and shipped apps used by 20,000+ people.

Data science student, ML practitioner, app builder
I'm Jeevan Koiri, from Nepal and now living in Milan, where I study for a Master's degree in Data Science at the University of Milano-Bicocca (UniMiB). Before that I earned a Bachelor's in Electronics, Communication & Information Engineering at Pulchowk Campus, IOE, Tribhuvan University, with coursework in AI, Applied Data Science, Big Data (Hadoop), DBMS and algorithms.
My data work started with computer vision: my final-year project trained YOLO to find people in thermal infrared images, placed in the Top 12 at the ICT Award 2023, and became a paper at the IEEE International Conference on ICT and Photonics 2026. As a Data Analyst & Software Developer at CG Electronics I built Power BI dashboards over 35,000+ monthly service calls, which helped cut average resolution time by 20%, and Python automation for 25,000+ inventory records. I also prototyped a concealed weapon detector and a satellite-imagery dashboard for algae blooms.
I'm also a software engineer who ships: 7 Flutter apps on Google Play and the App Store with 20,000+ downloads, plus production websites. That means I can take a model all the way to something people use. I'm looking for data science and machine learning internships and roles in Milan or remote.
Models, dashboards and analysis I've built
A low-cost, portable detector that finds people in darkness, smoke or foliage: a YOLOv3 model trained on thermal images, running on a Raspberry Pi 4B with a FLIR Lepton 3.5 camera and GPS. It reached an AP50 of 0.94, and a Flutter app sends real-time alerts with location and image. Published at the IEEE International Conference on ICT and Photonics 2026.
Power BI dashboards tracking 35,000+ service calls a month across spare-parts operations. Gave managers real-time visibility and helped reduce average call resolution time by 20%. Also automated pricing entry for 25,000+ inventory items with Python and Selenium, cutting manual work by 40%.
YOLO models trained on concealed-pistol and thermal imaging datasets to flag hidden weapons in images, with a Streamlit app for live demos and evaluation.
Analysed satellite imagery with Google Earth Engine and Python to estimate algae blooms and turbidity in lakes, and built an interactive dashboard for monitoring and decision support.
Generates PowerPoint presentations from long text using the Pegasus transformer for abstractive summarisation, with a Django backend running the pipeline.
Production software I've built and shipped

Coaching and chat companion app for the Afriqueen dating platform, built solo for iOS and Android. Reached 10K+ downloads within its first week and holds a 4.9 rating on Google Play. A dating app for the same client is launching soon.
































Exam-prep app for CTEVT students in Nepal: notes, past questions, syllabus and results. 10K+ downloads on Google Play. Built from scratch and still maintained by me, with BLoC architecture, a Firebase backend, OAuth 2.0 and an admin panel.

Wellness booking app on the App Store. On the team I owned the AI chat feature: integrating the Django backend's API endpoints and matching the Figma design pixel for pixel in Flutter.

Full-stack website for a Kathmandu trekking company: travel packages, trip planning and booking enquiries. Built with React and Next.js; I designed, built and still maintain it.

Website and gym management system for Eagle Gym Center in Lalitpur, covering memberships, plans, opening hours and enquiries over WhatsApp.
The technologies I use to bring ideas to life
The milestones that shaped my career
Courses include Machine Learning and Decision Models, Data Management and Visualization, Statistical Modelling, Text Mining and Search, and Foundations of Computer Science. Alongside coursework I'm extending my thermal-infrared / RGB computer-vision research.
Built the Afriqueen Assistance app solo for iOS and Android: 10K+ downloads in its first week and a 4.9 rating on Google Play. A dating app for the same client is launching soon.
Data and ML work: built Power BI dashboards tracking 35K+ service calls a month, which cut average call resolution time by 20%; automated pricing entry for 25,000+ inventory items with Python and Selenium (40% less manual work); built a YOLO + Streamlit concealed-weapon detection prototype. At DroneWorx.ai, built a Next.js control app with ROS & Gazebo for an autonomous weed-cutting vehicle. Stack: Python (pandas, NumPy, seaborn, scikit-learn), PyTorch, Power BI, SQL, Excel.
Part of the team building the MyNewEarth wellness booking app (App Store). I owned the AI chat feature: integrated the Django backend's API endpoints (Swagger-documented) and matched the Figma design pixel for pixel. Stack: Nylo (Flutter), Django REST APIs, Firebase, Git.
Co-founded CTEVT Plus and still run it. Designed, built and released the app (10K+ downloads on Google Play) with a Firebase backend, OAuth 2.0 and an admin panel. Also built and deployed ktmacademy.com.
Created frontends for mobile application and website to facilitate hotel businesses in Nepal. Stack: Flutter, Figma, Git, Android SDK, APIs, MVC Architecture.
Co-author of "An Efficient Infravision Thermal-Infrared Human Detection System for Low-Visibility Surveillance" — a YOLOv3 thermal detector on Raspberry Pi with an AP50 of 0.94. DOI: 10.1109/ICTP67998.2026.11485207.
Major project "Detection and Locating of Humans using Thermal Infrared Images and YOLO" selected in Top 12 Innovation Category at ICT Award 2023.
Won the Electric Vehicles category of the "Two Minutes Video Challenge" organized by IEEE IoE Pulchowk student branch.
Percentage: 76.27% (First Division), class rank 11 of 48. Courses: AI, Applied Data Science (Samsung), DSA, Big Data Technologies (Hadoop), DBMS, Probability & Statistics, Linear Algebra. Certifications: AWS Data Engineering, Samsung AI, Harvard CS50.
Completed higher secondary education in Science stream.
Completed Secondary Education Examination with distinction.
Delivering end-to-end solutions from concept to deployment
Turning messy operational data into answers: cleaning and analysis in Python and SQL, and Power BI dashboards that teams check every day. My dashboards at CG Electronics covered 35K+ service calls a month.
See projectsTraining and deploying models: YOLO object detection on thermal and RGB images, transformer-based NLP, and putting models inside apps and Streamlit demos people can actually use.
Contact MeEnd-to-end mobile application development using Flutter and Firebase. From UI design to store release: I have shipped 7 apps with 20,000+ downloads between them.
View on Play StoreTaught 100+ students in Computer Science, Engineering Mathematics, and programming. Available online for secondary, +2 and bachelor level, in English, Nepali or Hindi.
TeacherOn | Prosikshya | Book a CallA structured approach to delivering quality results
Understanding your requirements, goals, and target audience through in-depth discussion.
Creating wireframes and prototypes that align with your vision and user experience best practices.
Building your product with clean, scalable code using modern frameworks and best practices.
Rigorous testing across devices and scenarios to ensure reliability and performance.
Deploying your product and providing documentation, support, and maintenance guidance.
What the people I have worked with say
Verified credentials from industry leaders
Amazon Web Services
Simplilearn
Samsung
Harvard University (edX)
Coursera
Bootcamp
Electronics, Communication & Information Engineering
My open source activity
Have a project in mind? Get in touch
jeevankoirima@gmail.com
+977 9807590455
Copyright © Jeevan Koiri