Tushar Gupta
🚀 Welcome to my Portfolio!
💻 I do MLOps and other cool stuff!
📜 Background
💡 I am a final (third) year undergrad majoring in Data Science at TU Eindhoven.
📕 I have done several courses in machine learning, natural language processing, probability theory, statistics, econometrics, data mining, data structures, logic, algorithms and so on.
🏃 My hobbies include gym, cooking and painting occasionally. My nerdier side is always on the lookout for amazing open source projects and github repos that make my life easier :)
⭐ My Favorite GitHub Repos Currently
🤖 Open Operator - open source alternative to OpenAI's Operator
🔍 Cloudy Pad - open source alternative to deploy your own server for cloud gaming
📊 Exo - unify iPhone, iPad, Android, Mac, NVIDIA, Raspberry Pi or pretty much any device to make one powerful GPU
🧠 mactop - terminal-based monitoring tool for macs with silicon chips
👩🏽🚀 My Side-Projects
📚 Tutor101 - RAG for Exam Preparation
📭 2 Minute Marketing - LoRA Stable Diffusion for Marketing for SMEs
🕵️ Analysing Trust & Confidence of Londoners towards MET Police
💭 A goal without a deadline is just a dream
-Robert Herjavec
📆 Work History
🔃 Student Assistant for Database Modelling at TU Eindhoven [Current]
- Developing an educational app for practicing & learning SQL with Prof. Nick Yakovets & Dr. Hildo Bijl.
- Focusing on backend to implement the database logic in the local storage and Firebase for hosting with Vite and React for the frontend.
- Responsible for creating SQL exercises focusing on relational databases & various SQL functions, following a tree learning structure.
⚡️ Undergraduate AI/ML Researcher at TU Eindhoven [Current]
- Developing AI agents to enhance business model canvas’ creation for entrepreneurs.
- Containerizing agents using Docker and deploying via Vercel with NodeJS with KV database integration.
- Collaborating with researchers and faculty to scale applications across the BeNeLux region.
🩻 Image Processing Intern at High Tech XL - INNER
- Optimized image processing models for anomaly detection in car battery packs.
- Utilized Anomalib library (OpenVINO, PyTorch) in Python for implementing algorithms like padim, patchcore, and ganomaly for image segmentation.
- Implemented Docker containerization for efficient deployment in a controlled environment.
📈 Data Analyst at FLUOR Corp.
- Designed an interactive dashboard using time-series forecasting for salary prediction of 5,000+ employees over the upcoming year.
- Used SQLite for combining different schemas in the databse which was then implemented to fit a linear regression model to produce interpretable results.
- Created visualizations with Plotly to visualise the interpretable results and dynamic updates with new data.
🙏 Thanks!
📱Connect with me on LinkedIn or check out my projects on GitHub.