Kushal Pandya

AI/ML Enginner

AI/ML Engineer · Kody TechnolabCharotar University of Science and Technologyfull_time, internship

Open to roles

Machine LearningPython

Student Expense Tracker

This is a Python command-line Student Expense Tracker. It runs with the Python standard library only, with no database or web framework required. How to run: python student_expense_tracker.py The program includes 3 sample expenses on first run, so reviewers can immediately view expenses, total spending, and highest spending category. It also includes the bonus monthly budget warning feature. Expenses and budget are saved locally in expenses_data.json so data remains available after closing the program.

5 media files · github.comView

MemoryVerse AI '26

MemoryVerse AI is a fully local, zero-cost AI system. No paid API keys required. Setup to test locally: 1. Install Ollama -> run: ollama pull mistral 2. Download Qdrant binary -> run: qdrant.exe 3. pip install -r requirements.txt 4. uvicorn backend.main:app --reload --port 8000 5. cd frontend -> npm install -> npm start 6. Open http://localhost:3000 Demo user_id: u_test_01 (pre-loaded in the UI) Key things to evaluate: - Upload any PDF/DOCX on the Upload page -> watch live processing log - Ask natural language questions on Dashboard -> see cited answers with confidence scores - Click Knowledge Graph -> Rebuild Graph -> click any node to see connections - Gap Analyst -> select "ML Engineer" -> see skill radar + learning roadmap - Interview Coach -> generate Technical question -> submit answer -> get STAR feedback All 5 hackathon modules implemented + 2 beyond-the-brief features. RAG pipeline uses hybrid search (Dense + BM25 + RRF fusion) -> visible in search result scores.

5 media files · drive.google.comView

IPL_match_analytics

Analyzes IPL data to find what drives T20 wins: toss, phases, wickets and dot-ball pressure. Process Cleaned IPL data, engineered features, performed EDA, and visualized match insights. Results Middle overs, wickets, and dot balls impacted wins more than toss advantage. Reflection Build an interactive dashboard and add predictive modeling for future match analysis.

6 media files · drive.google.comView

This is Kushal’s work on Wooble.

Build a profile that shows what you can do — and share it anywhere.

Build yours