Resume

RAJAT PATRA

+91-7327047778|rajatkk.patra@gmail.com| github.com/RawJat| linkedin.com/in/patrarajat

Education

Bachelor of Technology in Computer Science and Engineering

2022-2026
Siksha 'O' Anusandhan University, BhubaneswarCGPA: 8.36

Projects

Cricket Fantasy XI Predictor | Python, XGBoost, PuLP, Docker

  • Developed an AI/ML-based predictive model to optimize fantasy cricket team selection for the Indian Premier League 2025, leveraging historical player data, pitch conditions, and weather to maximize fantasy points across 15+ matches.
  • Implemented Gradient Boosted Regressor and constraint optimization algorithms using PuLP to ensure adherence to fantasy sports rules, including a 100-credit budget and team composition constraints.
  • Processed and analyzed large-scale cricket ball-by-ball datasets from Cricsheet, Kaggle, and ESPNcricinfo, performing feature engineering to create batting and bowling feature sets for player performance prediction.

Envault | Next.js 14, Supabase, Redis, AES-256-GCM, Tailwind CSS

  • Architected a zero-knowledge vault using envelope encryption (AES-256-GCM) with a master/data key hierarchy and automatic rotation. Optimized decryption performance via Redis caching and enforced database security using Supabase RLS and immutable audit logs.
  • Developed a CLI for secret synchronization, engineering automated safeguards like git-tracked file blocking and pre-commit hooks to prevent credential leaks. Implemented the RBAC and Just-in-Time (JIT) GitHub integration to dynamically provision access to repository collaborators.

AI Generated Image Classification | PyTorch, Torchvision, scikit-learn, imbalanced-learn, Matplotlib

  • Worked on a Vision Transformer (ViT-B/16) model fine-tuned on the CIFAKE dataset (100k+ images) to classify real vs AI-generated images.
  • Achieved 98.5% accuracy and a 98.2% F1-score in the standard test set and maintained robustness against adversarial examples.
  • Optimized training with GPU acceleration (CUDA), reducing training time by 85%.

HealSync | Python, Figma, Adobe

  • Designed the application interface using Figma and Adobe Photoshop, which was later developed in Android Studio.
  • Built an RFID-based system in Python to track hospital facilities availability in real time.
  • Integrated Google Maps API to discover nearby hospital facilities to users.

Freelance Experience

Ekaayam | Nodejs, React, Flask, Firebase API, Razorpay API

  • Engineered a ticketing platform that reduces manual booking effort by 70%, enabling seamless booking options.
  • Redesigned and optimized the website to align with the brand identity, resulting in a 79% increase in website clicks.

Technical Skills

  • Programming: Java, Python, C, C++, VHDL
  • Development: HTML, CSS, JavaScript, Typescript, React, NextJs, Tailwind, MySQL, PostgreSQL, Supabase
  • Data Science: Numpy, Pandas, Matplotlib, Scikit-learn, TensorFlow, Statsmodels, Keras, PyTorch
  • Other tools: Kaggle, Git, Jupyter, Figma, Canva, Adobe Photoshop

Positions of Responsibility

Google Developer Student Club

  • Organised and managed the backend operations for Techwiz 2024, an event presented by Coding Ninjas and GDG ITER, attended by 200+ students.
  • Served as a key anchor for the Google Solution Challenge 2025, along with Google Tech Winter Break.

Academic Achievements

  • Secured First Prize (Rs. 10 Lakhs) in Dream11 Gameathon 2.0, competing against 40+ teams from Tier-1 universities across India, as part of a Rs. 25 Lakh prize pool.
  • Finished 7th in Pravaah ML Hackathon, organized by IIT Bhubaneswar.
  • Achieved 1st Rank in the Science Olympiad organized by the Science Olympiad Foundation and also qualified for the state level.
  • Qualified in the internals of the Smart India Hackathon 2024 conducted by the college with a rank of 11 out of 800+ participants.