I build software across the stack — from Android applications and self-hosted deployment platforms to research-grade machine learning pipelines. My current focus is on deepening my understanding of machine learning: representation learning, multi-objective optimisation, and applied NLP with deep learning. I am actively looking to learn more in this space through research, coursework, and open-source work.

A productivity and focus application for Android and iOS.

20,000+ downloads across 25+ countries · 1,500+ daily active users

View Goal Guard on Google Play

The project I am most proud of. An open-source coordination and context layer for AI coding agents — a shared, versioned memory that follows you across Claude Code, Cursor, Copilot, Codex, and Antigravity, so switching tools no longer means losing the plot.

Node.js · TypeScript · Model Context Protocol · Next.js · Multi-Agent Systems

View the repository on GitHub

Languages

Python · C++ · C · Rust · Go · Java · Kotlin · JavaScript · TypeScript · SQL · CUDA · Verilog

Deep Learning and Artificial Intelligence

PyTorch · TensorFlow · Hugging Face Transformers · Accelerate · PEFT and LoRA · ONNX · OpenCV · YOLO · Optuna

Generative AI, Agents, and the MCP Ecosystem

Model Context Protocol (MCP) · MCP servers over stdio and HTTP · Tool-calling and function-calling agents · Multi-agent orchestration · Agent memory and context engineering · Claude Code, Cursor, Copilot, Codex, and Antigravity integrations · LangChain · LangGraph · LlamaIndex · Retrieval-Augmented Generation · Vector search and BM25 hybrid retrieval · Prompt and context compression · Fine-tuning and evaluation harnesses · vLLM and Ollama · OpenAI and Anthropic APIs

Classical Machine Learning and Data Science

Scikit-Learn · XGBoost · LightGBM · Pandas · NumPy · SciPy · Matplotlib · Genetic algorithms and multi-objective optimisation

Backend and Application Development

Node.js · FastAPI · Flask · Spring Boot · React · Next.js · Tailwind CSS · Android SDK · Jetpack Compose · Unity

DevOps and Infrastructure

Docker · Kubernetes · Caddy · Git and GitHub Actions · Linux · Google Cloud Platform · PostgreSQL · SQLite · Redis

• Machine learning and deep learning — I am currently learning and building in this area, with coursework in Machine Learning, Multi-Objective Machine Learning, Natural Language Processing with Deep Learning, Machine Reasoning, and Network-Based Computing for HPC and AI.
• Applied research in physiological signal processing and model efficiency for edge deployment.
• Agentic systems, LLM context compression, and developer tooling.
• Competitive programming and algorithm design.

Contributions to my open-source work are genuinely welcome. RelayBrain in particular is built to be extended, and new agent integrations are the highest-value contribution right now.

1. Open an issue describing the bug, integration, or feature before starting substantial work.
2. Fork the repository and create a branch named for the change, for example feat/cursor-adapter.
3. Keep commits focused and write commit messages that explain the reasoning, not only the diff.
4. Ensure the existing checks pass, and add tests covering new behaviour.
5. Open a pull request that describes the problem, the approach, and any trade-offs you accepted.

Bug reports, documentation improvements, and reproducible test cases are as valuable as code.

I am open to research collaborations in machine learning, contributions to open-source developer tooling, and internship opportunities. If you are working on something in these areas, please reach out.