ML at Rippling

Building agent harnesses for Rippling AI.

Previously - Founding Applied Scientist @ Typeface.ai, Applied Scientist II @ Azure OpenAI & Microsoft’s AI Development Acceleration Program (MAIDAP) at the Office of Applied Research.

Over the years, my work has spanned the breadth of problems in the areas of Deep Learning and ML. In additition to NLP, I have previously worked on major projects in the domains of Vision, Interpretability, and ML for Systems.

Past Highlights:

Building Typeface’s text-gen stack (2022-25)

Built first-to-market product using LLMs for style-compliant content gen for enterprise marketing. Helped scale from stealth to Unicorn valuation. Built prompt engineering stack, content stylization stack, web-research stack, finetuning flows & early agent harnesses for the text-gen product.

Satya Nadella’s GPT-3 Ignite Keynote (2021-22)

Our work on using GPT-3 for video summarization and blog generation for WNBA was featured as one of 2 demos in Satya Nadella’s 2021 Ignite keynote. It was one of the first large demos showing the promise of free-form workflows with LLMs.

Optimal Resource Allocation for Big Data workloads (2020-2021)

Our work on exa-byte scale resource use prediction from computational graphs was published in EDBT 2022 with associated patent.

Professional Summary:

Publications:

• EDBT 2022 - Towards Optimal Resource Allocation for Big Data Analytics
• ER 2020 - Evaluating Tree Explanation Methods for Anomaly Reasoning
• US Patent - Optimizing job runtimes via prediction–based token allocation
• US Patent. - Intelligent table suggestion and conversion for text
• US Patent - Proactively generated content and personalized feeds available for audiences
• US Patent - Interactive prompting system for multimodal personalized content generation

Past Experience: 8 yrs

• Founding Applied Scientist at Typeface for 2.5 years
• Applied Scientist II at Microsoft for 3.5 years
• Deep Learning intern - Computer Vision at Mathworks for 8 months
• ML Engineer at SenseHawk for 6 months
• Executive Engineer - Data Analytics at Honda for 6 months

Education:

• MS in CS at UMass Amherst

Misc:

I spend most of my leisure time cooking, learning musical instruments and climbing (with varying degree of success).
I love ranting about terrible urban design and discussing geo-politics.
I am fascinated by all sorts of meta information, which doesn’t help my case of going down random internet rabbit holes.