What I Do

• LLM finetuning & custom deep learning: Model training, supervised and RL finetuning
• DataOps and MLOps: Building custom data and ML pipelines to terabyte (TB) scale and beyond
• AI agentic systems: Building agentic systems that can perform tasks autonomously
• AI SaaS: Backend, ML, Frontend, CI/CD, Deployment, Scaling
• Hardware-efficient bio/large models: Developing efficient architectures for biological sequence understanding — esmc.cpp, PepEmb

About me

I am a biomedical engineer turned machine learning (ML) engineer with experience building ML systems since 2018. I have worked at academic labs like TERI, IGIB, and IGBMC. My work revolved around building statistical and machine learning models in healthcare and biological sciences (genomics, proteomics, etc).

As a freelance consultant, I've worked in the patenting, e-commerce analytics, and SaaS industries.

Some of my work:

E-commerce: Systems for named entity extraction, information organization, product embeddings, information retrieval systems like semantic search, RAG, and low latency extreme classification tasks.

SaaS: Low latency (sub 100 ms) task-specific ML systems like classifiers for both image and text modalities, custom embedding models, and information extraction systems.

Patenting: Evaluated the feasibility of AI systems for patent-related applications.

For the past couple of years I have been learning full stack engineering. My tech stack of choice is:

• Pytorch
• FastAPI
• Vanilla JavaScript
• Postgres
• Docker
• Celery
• Nginx
• Git
• Coolify
• HTMX
• AWS
• Hetzner

I like to self host software. Currently I am self hosting coolify, databases, plausible, and my own websites and products.

Work

Currently I am working on building PatViz, which automates patent drawings and for patent researchers and agents.

I am also building DataEngUtils, a platform for data engineering utilities.

I am a maintainer of esmc.cpp, a C++ inference engine for protein language models.

Previously I was working on building AdTestPro (archived), which provides a platform for testing and optimizing ad campaigns.

Writing

I write about ML engineering, technology, and lessons learned from building production systems. Check out my blog for technical insights and project deep-dives.

Contact

If you have a project in mind or want to explore ideas, reach out.