Data Scientist · Deep Learning · NLP · MLOps
I build ML systems end-to-end — from raw data to deployed models. I believe the best way to understand something is to build it from scratch.
I'm a Data Scientist based in India with a focus on building ML systems that work in production — not just in notebooks. My work spans the full lifecycle: data pipelines, model development, and deployment.
I write on Medium about neural networks, LLMs, and applied ML — always from first principles. If I can't build it from scratch, I don't feel like I truly understand it.
Currently interested in: production RAG systems, LLM fine-tuning, and scalable ML pipelines on Databricks.
First-principles ML → Backprop, gradient descent, perceptrons — built by hand
Applied LLMs & RAG → LangChain, LlamaIndex, vector DBs, production pipelines
Data at scale → PySpark, Databricks, pipelines for large datasets
MLOps → FastAPI, Docker, model serving and monitoring
| Area | Tools |
|---|---|
| Languages | Python · SQL · PySpark |
| ML / DL | PyTorch · TensorFlow · Keras · Scikit-learn |
| NLP | SpaCy · HuggingFace Transformers |
| LLM & RAG | LangChain · LlamaIndex · Vector DBs |
| Data | Pandas · NumPy · Databricks |
| Infra | FastAPI · Docker · Git |
I write about ML concepts by building them from scratch. No hand-waving — just code and math.
| Article | Topic |
|---|---|
| Creating a RAG application from scratch | Retrieval-augmented generation, end-to-end |
| Gradient descent from scratch | Optimization fundamentals |
| Backpropagation in neural networks from scratch | How neural networks actually learn |
| Training a single perceptron from scratch | Where deep learning begins |
Open to collaborations in applied ML and LLMs · Based in India