SG Shubham Goel

About

Hi, I'm Shubham.

I'm a machine learning engineer with computer-vision roots. Most recently I was a Senior MLE at Meta on Instagram Ads Ranking, where I worked on multimodal LLM ranking features, model distillation, and the eval harnesses that keep all of it honest.

Before Meta, I spent a few years at ZEFR building multimodal search and content-understanding systems, did my M.S. in Statistics at UIUC, and worked across computer-vision roles at Reliance Jio and Pitney Bowes, spanning satellite imagery, building detection, and more.

I firmly believe that specialization is for insects, which is why I keep a few too many interests running at once: technology, startups, education, philosophy, and an unshakeable allegiance to Arsenal.

Shubham Goel at the beach, smiling

The path here

  1. Senior Machine Learning Engineer

    2024–2026

    Meta · Instagram Ads Ranking

    • Built multimodal LLM pipelines generating ad-side ranking features, and post-trained open models (SFT, teacher-student distillation) against custom per-field eval harnesses.
    • Shipped a domain-tuned video OCR pipeline reaching near-parity with a frontier VLM at a fraction of the cost.
    • First author on a RecSys '26 Industry Track submission on text-on-creative as a ranking modality.

  2. Senior Machine Learning Scientist

    2021–2024

    ZEFR

    • Built a multimodal semantic search platform over 1B+ image and text embeddings, powering CV/NLP fine-tunes across brand-safety taxonomies.
    • Fine-tuned CLIP/ViT models and built a vector-search active-learning loop. Invited speaker at the Snap ML group (2023).

  3. M.S. in Statistics

    2019–2020

    University of Illinois Urbana-Champaign

    • GPA 4.0. Research on hurricane property-damage estimation from satellite imagery and insurance data.

  4. Machine Learning Engineer, Computer Vision

    2018–2019

    Reliance Jio

    • Building-footprint extraction at scale to automate population estimation for the JioFiber rollout.

  5. Software Engineer, ML

    2016–2018

    Pitney Bowes

    • Where the machine-learning thread of my career started.

Education

University of Illinois Urbana-Champaign, M.S. in Statistics (GPA 4.0)

University of Delhi, B.Tech, I.T. and Mathematical Innovations