LOCATIONS: STOCKHOLM | DUBAI
Session Outline

End to end system design which abstracts out different processes in a typical ML project. Hyper configurable system governing the 3 main processes of ML project – Data Pipelines, Model learning and end consumption.

Key Takeaways

  • Building a self serve AI engine at an enterprise level
  • Enabling 4x efficiency both in terms of time and cost to delivery of ML models
  • Module level system design which can enable non-tech folks deploy ML models at scale in a seamless fashion

————————————————————————————————————————————————————

Bio

Ankur Verma – Senior Data Scientist | Amazon India | India

I have more than 9 years of industrial experience in the field of AI/ML. Currently I am working as Senior Data Scientist at Amazon India where I lead a team of ML scientists. My responsibilities here at Amazon include writing end to end MLOps from creating data pipelines to building scalable ML/DL models using state-of-the art engineering stack. Most of my experience has been solving AI problems for large retail firms like Walmart, Amazon, Dunzo, etc. Lately, since the past 3 years, I have been keenly invested in designing systems for solving ML problems in a platform based fashion at an enterprise level. Demand Forecasting and Offer Personalisation using deep net architectures in a hyper configurable manner are 2 main highlights of the problems I have been involved in the last couple of years.

January 12 @ 11:35
11:35 — 12:05 (30′)

Stage 2

Ankur Verma – Senior Data Scientist | Amazon India | India