Jan 2023 – Present · 3 yrs 5 mos

• MLops platform backed by Sequoia, Naval Ravikant and Anthony Goldbloom (Kaggle)

Jun 2022 – Present · 4 yrs

Oct 2021 – Present · 4 yrs 8 mos

Jul 2021 – Present · 4 yrs 11 mos · San Francisco Bay Area

• Helping advise in graph and NLP technologies to build effective solutions for extracting and insights from text and unstructured data.

Feb 2020 – Present · 6 yrs 4 mos

Feb 2020 – Sep 2021 · 1 yr 7 mos · Bengaluru, Karnataka, India

• Technical Advisor on Machine Learning and Artificial Intelligence.
• Company was acquired by Fractal

Feb 2018 – Aug 2020 · 2 yrs 6 mos · San Francisco Bay Area

• This group was led by Prof. Zoubin Ghahramani
• ( https://en.wikipedia.org/wiki/Zoubin_Ghahramani).
• I am involved in applied AI research on Uber datasets and have deployed multiple models at scale at Uber. Specifically working(worked) on:
• + Large Scale Graph Representation Learning algorithm for Uber Eats, Fraud detection, Traffic
• Prediction. This application mainly involves usage of Graph Convolutional Networks.
• + Food Preparation Time Prediction in Uber Eats using Deep Neural Networks.
• Tools: Python, R,Pytorch, Tensorflow, Hive, Docker, PySpark
• Additionally, I run the internal AI education initiative at Uber. This has helped hundreds of engineers at Uber to learn and use AI in their jobs.

Mar 2017 – Jan 2018 · 10 mos · San Francisco Bay Area

• User Level Forecasting (Advised by Chief Scientist, AI Labs):
• Led the effort on predicting # trips for each rider/driver on the platform.
• o Developed a LSTM model with custom loss function (Zero Inflated Poisson) using PyTorch to predict trips for each individual driver in the short term (4-6 weeks). Incorporated incentive features and developed first of a kind incentive sensitivity curves for the company at aggregate level
• Self Driving Car Simulation: Serving as a technical lead for developing an agent based simulation using Python to help business decide on the strategy for deploying self driving cars
• Tools Used: Python, R,Pytorch, Tensorflow, Hive, Machine Learning, Deep Learning, BI Tools

Sep 2016 – Mar 2017 · 6 mos

• Headed core technology and worked on developing patent technology on Inpainting of Home Improvement objects through Computer Vision.
• Company was part of Nvidia Inception program and had provisional patent applied. .
• Applied for provisional patent: “SYSTEM AND METHOD FOR IDENTIFYING, RECOMMENDING AND REPLACING VISUALLY SIMILAR PRODUCTS IN THE HOME IMPROVEMENT DOMAIN”
• Drishyam AI got acquired by Mediaocean in March 2022!!

Oct 2015 – Dec 2016 · 1 yr 2 mos · Bengaluru Area, India

• Runnr was acquired by Zomato.
• Solved complex last mile logistics problems in India with a data twist. Here is a brief of my work:
• Led a team of 5 data scientists/analysts and 4 engineers while being an individual contributor.
• Food Preparation Time (FPT) Prediction
• Analyzed the data and built a Random Forest Regressor using scikit-learn to predict food
• preparation time based on hour, day, cost of order etc.
• Model was used to optimally dispatch the delivery boys and achieved 7% increase in efficiency
• Grouping of Ecommerce Delivery: Developed a custom model on the lines of hierarchical clustering to group the ecommerce deliveries. Model led to 5% increase in number of orders delivered with the same fleet and similar SLA.
• Driver Churn Prediction Model: Developed a churn prediction model to preempt the driver churn using Logistic Regression. Operationalized the model which led to 15% decrease in churn of drivers
• Led several other data science projects using Time Series Analysis like Demand/Supply Prediction etc.
• Tools Used: Python, R, BI Tools, SQL, Redshift SQL, Django, MongoDB.
• Here is the link to my statistical analysis on Food delivery business in India.
• http://yourstory.com/2016/07/food-delivery-sustainable-business/

May 2015 – Dec 2017 · 2 yrs 7 mos · San Francisco Bay Area

• Mentor for Data Science course. Making students fall in love with data.

Mar 2014 – Sep 2015 · 1 yr 6 mos · San Francisco Bay Area

• Worked on problems relating to prediction of probability, time and amount of a sales deal close
• Tools Used: MySQL, Redshift, R, Python, Tableau. Filed a patent on the same.

Apr 2013 – Feb 2014 · 10 mos · San Francisco Bay Area

• Working in the Capital Management Group. Analytics using Teradata SQL and Statistical Softwares.

Jan 2013 – Mar 2013 · 2 mos · Menlo Park CA

• Worked in product analytics team on Big data projects.
• Extensively programmed in Hive, SQL, R, Python.
• Machine Learning, Feature Engineering.

Feb 2009 – Feb 2012 · 3 yrs · Mumbai Area, India

• Supervised the entire drilling operations worth $ 8 million on the rig sites by ensuring optimum utilization of resources and time.
• Led and trained cross-cultural teams of 4-7 engineers involved in the entire drilling operation on the rig site.
• Supervised and drilled the first ever offshore High Temperature and High Pressure well in Eastern India. Unlocked the future market for the same, worth $ 15 million.