About me

Hello! I am Romit. I love building data-driven tools and products. I have (Current year - 2018) years of work experience in making companies data-driven and building production-level ML/DL models.

A bit about my experience

• [December 2023-Present]
  • I am currently working at Meraki Labs as an AI engineer where I am working on synthetic data generation, RAGs, small-scale multimodal agents, and GPU optimizations
  • I primarily work with Triton, CUDA C, and PyTorch
• [November 2020-October 2023]
  • I worked at epiFi as a Senior data scientist where I built real-time computer vision and machine learning models
  • This involved building infrastructure for real-time and batch models, like a ton of real-time feature engineering (>1k features), optimizations for deploying deep learning and ML models with lower latency (TF lite, quantization, in memory DBs), and building highly observable micro-services
  • I worked extensively with Tensorflow, Docker, PySpark, Airflow, EFK stack, and the usual AWS stack
• [January 2018-November 2020]
  • I worked at Postman as a Technical lead in the data science team and was the founding member of the data team
  • I built and designed several internal tools including
    • Data APIs: This was my major work at Postman that helped Postman to be data-aware and data-driven
    • Data testing framework (Based on Looker and DBT)
    • Knowledge Repository
    • ETL pipelines
  • I worked on SQL, Redshift, Looker, DBT and PowerBI
• I was on the team that built the project Panoptic.

A bit about what I love doing

• Tech I love working with - Python, CUDA C, PyTorch, SQL, Docker, FastAPI
• Areas I am interested in
  • Applied deep learning, especially computer vision
  • Human in-loop decisions
  • MLOps
• Reading books, playing cricket, and petting cats
• Obsessed with calendars, to-dos, tracking every litte thing, and shiny productivity apps