Sanket Gupta

About Sanket Gupta

I am a Staff Machine Learning Engineer at Meta, specializing in large-scale sequence modeling and ranking systems within the Monetization domain. I build and evolve the ML engines that power high-stakes consumer platforms, operating across billions of daily impressions.Previously at Spotify, I led core personalization initiatives initiatives within the Home experience — driving measurable gains in retention, revenue, and ranking efficiency through next-generation user representations and foundational embeddings.My work sits at the boundary of research and production: designing novel modeling approaches while ensuring they scale reliably in real-world systems. I focus on high-leverage ML problems where model architecture, experimentation, and business impact intersect. Impact: a) Drove the development of a new “sequence-as-labels” modeling paradigm to go beyond sequence as features and use them for labels. b) Saved $2.1M annually at Spotify by successfully implementing transfer learning on user’s artist preference model. c) Helped build and scale next gen user representations deployed across multiple core models at Spotify. Published this work at RecSys 2025. d) Built and scaled user calibration models to decide mix of podcasts and music on Spotify Homepage and had multiple rollouts with DAU and WAU increases. Presented this in RecSys 2025 workshop. e) Spoke at Google Next 2025 on solving real time AI challenges at Spotify, and at MLConf 2024.

Experience

1. Staff Machine Learning Engineer

   Meta

   Sep 2025 — Present · New York, NY, US

   Working on sequence modeling and large-scale ranking models within Monetization, tackling some of the most challenging and high-impact ML problems at Meta.Projects: a) Helping develop temporal sequences of features on interactions and impressions for ranking models. b) Created a new pillar on using “sequence as labels” - going beyond traditional history oriented view of user activity and going beyond to using sequence for labels. c) Helping develop and champion an AI agent for new signal discovery and prioritization - helping save hundreds of hours of manual work and helping innovate new ideas.

Education

• Columbia University

  Master of Science (MS), Electrical Engineering

  2014 — 2015

• Nanyang Technological University Singapore

  Bachelor of Engineering (B.E.), Electrical and Electronics Engineering

  2007 — 2011

Skills

• Altium Designer
• Lvs
• Cadence
• Ic
• D3.js
• Vlsi
• Soc
• Analog
• Power Management
• Spice
• Mixed Signal
• Circuit Design
• Spectre
• Python
• Amplifiers
• Vco
• Integrated Circuit Design
• Asic
• Pll
• Analog Circuit Design
• Machine Learning
• Comparators
• Cmos
• Integrated Circuits (Ic)
• Semiconductors
• Verilog-a
• Physical Design
• Web Development
• Sql
• Microelectronics
• Cadence Virtuoso
• Social Media
• Bicmos
• Debugging
• Elasticsearch
• Arduino
• Linear Regulators
• Matlab
• Low-Power Design
• Internet of Things