Type: Full-stack, production-style system
Impact: Disaster awareness & preparedness

Key Features

• Multilingual disaster education modules (EN / HI / ML / KN)
• JWT authentication with role-based admin dashboard
• User progress tracking & analytics
• Mobile-first UX with Framer Motion

Tech Stack React · Node.js · Express · MongoDB

• 4 game modes
• JavaFX + MVC
• Live Java/Python execution via JDoodle API
• Supabase leaderboard + offline JSON caching

• Interactive heatmaps (Leaflet.js)
• Bengaluru-level emission simulation
• UX-first sustainability nudges

ETNA — Rust + Python ML Framework

Key Contributions

• Designed and implemented comprehensive pytest test coverage for preprocessing and utility modules, validating:
  • Numeric scaling
  • Missing-value handling
  • Categorical encoding
  • CSV ingestion
• Resolved Python-only test isolation issues in a Rust-backed ML framework, improving CI reliability.
• Fixed a critical model persistence bug by implementing save/load support for preprocessing state in ETNA’s Python API.
  • Ensured trained models retain scaling statistics, categorical encodings, and label mappings after reload.
  • Restored prediction consistency across sessions and validated correctness with end-to-end tests.
• Added native categorical feature support to the ETNA Preprocessor:
  • Automatic detection of string/object columns
  • One-Hot Encoding with robust NaN handling (mean for numeric, mode/unknown for categorical)
  • Dynamic feature-dimension updates for correct Rust core initialization
  • Guaranteed encoding consistency during inference
• Strengthened Python–Rust integration through unit tests and cross-language validation.

Tech: Python, Pytest, ML Pipelines, Feature Engineering, Rust–Python Interop

🔗 PRs:

• Pytest-based preprocessing coverage
• Preprocessor persistence (save & restore state)
• Categorical feature support with One-Hot Encoding

SoulSense — Open Source Developer

Key Contributions

• Refactored a monolithic Python EQ assessment tool into a modular, production-ready architecture with a clean package structure.
• Designed and implemented normalized, ML-ready SQLite schemas for:
  • Question banks
  • User responses
  • Question-level metadata
• Enabled fine-grained response storage to support analytics and future ML pipelines.
• Introduced structured logging and error traceability, replacing print-based debugging and improving observability.
• Implemented age-based question filtering while maintaining backward compatibility.
• Built a comprehensive pytest test suite covering:
  • Core logic
  • Database migrations
  • Utility modules
• Ensured migration-safe schema evolution and non-breaking enhancements across releases.

Tech: Python, SQLite, Database Design, Logging & Observability, Pytest, Modular Systems

🔗 Contribution summary & PRs linked in repo

CarbonX — Web3 Contributor

Key Contributions

• Built and integrated the GreenProof smart contract using Solidity and Remix to record eco-friendly actions on-chain.
• Ensured contract correctness and deployability as part of the CarbonX Web3 pipeline.
• Integrated GitHub Actions from the Marketplace to automate repository workflows:
  • Copying issue labels to PRs
  • Assigning PR authors automatically
  • Managing stale issues

Tech: Solidity, Smart Contracts, Remix IDE, GitHub Actions

🔗 PRs:

• GreenProof smart contract
• GitHub Actions workflow automation

AlgoVisualizer — Maintainer & Open Source Developer

Key Contributions

• Maintained a large-scale educational platform during Hacktoberfest & GSSoC '25:
  • Issue triage
  • PR reviews
  • Contributor mentoring
  • Merge conflict resolution
• Designed and implemented interactive visualizers:
  • Dynamic Programming (LCS, Coin Change, Matrix Chain)
  • Hashing (insert/search/delete with animations)
• Implemented advanced sorting algorithms with full animations:
  • TimSort
  • IntroSort
• Restored missing algorithms after a major refactor and modularized the codebase for long-term maintainability.
• Fixed a critical production build failure (Home.jsx), unblocking 30+ contributors.
• Expanded the quiz platform with 80+ algorithmic questions across:
  • DP, Backtracking, Divide & Conquer
  • Trees, Knapsack, TSP
  • Game Search (Minimax, Alpha-Beta, MCTS)
• Improved UI/UX clarity and documentation to enhance accessibility and learning outcomes.

Tech: JavaScript, React, Data Structures & Algorithms, Visualization, OSS Maintenance

🔗 Representative PRs:

• DP Visualizer
• TimSort & IntroSort animations
• Quiz platform expansion
• Production build fix

Old-School-Game — Open Source Contributor

Key Contributions

• Built a time-based arithmetic game with multi-level difficulty (Amateur → Normal → Veteran).
• Implemented dynamic scoring, attempt tracking, and instant validation.
• Designed Veteran-mode hints using mental-math techniques.
• Enhanced UI with responsive layouts, accessible inputs, and keyboard navigation.
• Incorporated CodeRabbit feedback to improve accessibility and styling.
• Proposed and implemented Bullet Math, a fast-paced cognitive game backed by research on brain training.

Tech: TypeScript, React, Accessibility, Game Design, UI/UX

🔗 PRs & proposal links in repo

PyEveryday — Open Source Contributor

Key Contributions

• Refactored and expanded the Data Converter Toolkit using pandas-based pipelines.
• Reduced codebase size by ~35% while improving readability and maintainability.
• Added preview functionality, row/column statistics, and data sanitization.
• Proposed and designed a modular ingestion architecture:
  • DataReader
  • DataProcessor
  • DataConverter
  • DataWriter
• Identified duplicated chunking logic and proposed unified read methods for consistent data handling.
• Collaborated with maintainers and contributors under OSCI ’25.

Tech: Python, pandas, Data Engineering, Modular Architecture

🔗 PRs & architectural proposals linked in repo