I am a computer scientist and engineer focused on finding and solving meaningful non-trivial problems at the intersection of research and application. I love building things from zero to one and prefer working with small groups of likeminded mission-driven people who care deeply about the societal impact of their work.

Background

I have extensive experience building and scaling end-to-end machine learning models.

Currently, I am working on Large Language Models at Shopify as a Staff Machine Learning Engineer.

Previously, I was a founding member of the Artificial Intelligence group at HubSpot and helped grow it from 5 to XX+ members.

Here's a list of some of the things I've worked on.

  • Building LLM (Llama2, Falcon, Mistral, etc) fine-tuning, evaluation, and serving layers for our ML Ops platform

  • Fine-tuning LLM models for specific use case using SFT and DPO and serving them with vLLM

  • Real-time streaming anomaly detection and forecasting on time series data

  • Bayesian Active Learning

  • Markov chains for user action prediction

  • Text classification, clustering, and similarity

  • Search: semantic retrieval, ranking, and performance monitoring

  • Contextual multi-armed bandits for adaptive A/B testing

  • Recommender systems for lead scoring, suggested academy lessons, and search result ranking

  • Abuse detection for email lists

I studied computer science and business at Rensselaer Polytechnic Institute and spent most of my time competing in hackathons and doing some research on free-space optical communication (FSO). I then went to Georgia Tech and obtained a masters in computer science with a concentration in machine learning.

In my free time, I enjoy reading, writing, sailing, using Notion as my second brain, and engaging with all things automotiveβ€”from driving and working on cars to discussing the latest trends. I also dabble a bit in filmmaking and photography.

Select writing

Talks

Contact

You can find me on Twitter @MukulSurajiwale, as therealmukul on Github, or here on LinkedIn.

Sometimes I also write out my thoughts in long-form, which can be found on my blog.

I am based in the NYC and I love meeting new people. So if you want to grab a coffee, I'd more likely than not, be down.