PhD Mathematician → AI/ML Engineer
I build end-to-end machine learning systems grounded in rigorous mathematical thinking.
Thesis submitted (Topological Dynamics & Ergodic Theory, BITS Pilani). Actively transitioning to ML/AI Engineering.
🌐 Portfolio | 💼 LinkedIn | ✉️ Email
- Completing the ISL Translator — real-time speech-to-sign-language (FastAPI, Whisper, WebSockets, React)
- Deepening MLOps skills: model monitoring, CI/CD pipelines, containerisation
- Applying for ML/AI Engineering roles
| Project | What it does | Stack |
|---|---|---|
| Multi-Agent Math Content Bot (private) | Dual-LLM generator-critic pipeline (Llama 3.3 70B + DeepSeek-R1) with HITL Telegram review and automated 3×/day GitHub Actions scheduling | Python, Groq API, Telegram Bot API, GitHub Actions |
| Placement Prediction System | End-to-end MLOps pipeline: ingestion → training → serialisation → REST API → production deployment | Python, Scikit-learn, Flask, Joblib, Gunicorn, Render |
| AI Book Scanner Automation | Automated pipeline: Telegram photo → Gemini Vision → structured JSON → Google Sheets | Make.com, Gemini Vision API, Telegram Bot API |
| ML Projects Collection | Regression, classification, clustering, CNNs — Seoul Bikes, MAGIC Telescope, MNIST, Seeds | PyTorch, TensorFlow, Scikit-learn, Pandas, NumPy |
ML/DL
Backend & APIs
CI/CD & Deployment
Tools
- TensorFlow Developer Professional Certificate — DeepLearning.AI, 2026
- Google AI Professional Certificate — Google, 2026
- Machine Learning in Production — DeepLearning.AI, 2025
- Mathematics for Machine Learning and Data Science — DeepLearning.AI, 2025
- k-type chaos of Z^d-actions
- k-type entropy of Z^d-actions
- k-type chaos for induced group actions on hyperspaces (preprint)
- Student epistemological beliefs and learning attitudes in STEM education in a work-integrated learning setting
To teach is to understand. To learn is to question. To think is to connect.
Mathematics trains the mind to think clearly. AI allows those thoughts to shape the world.