🏔️Sumeet Ranka profile photo 4 on Tal
🏔️Sumeet Ranka profile photo 1 on Tal
🏔️Sumeet Ranka profile photo 2 on Tal
🏔️Sumeet Ranka profile photo 3 on Tal
🏔️Sumeet Ranka profile photo 4 on Tal
🏔️Sumeet Ranka profile photo 1 on Tal

🏔️Sumeet Ranka

Sidekick @ Shopify

India

9+ years of work experience

Shopify
Dec 20241 yr 9 mosPresent

India

Empowering merchants with AI based companions, their own Sidekick.
Microsoft
Sep 20231 yr 4 mosDec 2024

Vancouver, British Columbia, Canada

I contributed to a horizontal initiative to integrate unified embeddings for cross-product personalization, creating pipelines that generate 1M to 50M user profiles daily. I integrated LLM and RAG evaluation schemes, accelerating iteration speed and impacting over 400M users across three products. As Tech Lead, I also diversified and enhanced image recommendations using Bing and Edge signals, using LLM-based retrieval algorithms to boost retention and engagement by 7-9%, and reduced in-session and cross-session duplicates, cutting Total Time on Server by around 15%. I also integrated multi-modal feature embeddings, increasing Engaged Sessions per User by 2.21% and Daily Active Users by 3.03%, and developed scalable, modular code, reducing flighting time from 3-4 days to 0.5 day.
Microsoft
Jul 20212 yrs 3 mosSep 2023

Vancouver, British Columbia, Canada

Working in Bing Multimedia team at Microsoft.
Huawei
Feb 20201 yr 6 mosJul 2021

Toronto, Canada Area

Tools Used: C++, Python, Tensorflow, Tensorflow Lite , Android JNI, HiAi Platform, ONNX • Prototyped face detection and object detection for android systems. • Mentored four co-op intern students working on different ML inference projects • Contributor to drowsiness detection system, featured as a major product by Huawei CBG CEO • Consulted on multiple projects for architectural design and embedded systems related optimization.
Phenomic AI
May 20198 mosDec 2019

Toronto, Canada Area

Tools Used: Python, CellProfiler, AWS services The project aims at applying machine learning in the area of drug discovery. Over a thousand microscopy images get generated every week and eye-based analysis is tedious. Hence, to make the process efficient, the segmentation and the feature extraction process is automated using various AWS services and latest cellular segmentation algorithm. After the individual cellular area are identified and their features are extracted, it is followed by analysis to classify the cells in images into different cell types using suitable supervised or semi-supervised learning techniques. The analysis also involves validation and quantification of the observation made by biologists to understand the effect of the environment on which the cells are put to. --------------------------------------------------------------------------------------------------------- BACKGROUND Cell-cell interactions have been found to have a significant impact on drug efficacy. However, various experimental conditions are tested and manual analysis quickly became a bottleneck. My main aim is to analyze the microscopy images of multi-cell cultures and extract intensity, shape, and texture-based features using an automation pipeline. These features are then put for analysis to validate the observation made by the biologists or understand the effects of the environment on the cells in which they are put.
University of Toronto
Sep 20188 mosApr 2019

Toronto, Canada Area

For the course: Computer Networking Systems (Fall Term), Intro to Programming (Winter Term)
Samsung
Jun 20171 y 2 moAug 2018

Bengaluru

GitHub
@ranka47

🏔️Sumeet’s AI Stack

claude code
claude code
codex
codex

Education

University of Toronto

Master of Science in Applied Computing, Computer Science

University of Toronto

20182 yrs2020
Indian Institute of Technology, Guwahati

Bachelor’s Degree, Computer Science and Engineering

Indian Institute of Technology, Guwahati

20134 yrs2017

Join 🏔️Sumeet and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

🏔️Sumeet Ranka on Tal

Headline

Sidekick @ Shopify

About

Experienced Data and Applied Scientist at Microsoft with a strong background in machine learning, cross-product personalization, and scalable solutions. I have successfully led initiatives to integrate unified embeddings, enhancing user engagement and retention across multiple products. My expertise is adaptive to the technology stack I work on, with a proven track record in developing efficient, modular code and innovative ML models. Passionate about leveraging technology to drive impactful results and foster collaborative environments.

Location

India

AI stack

claude code

codex

Work experience

Applied ML Engineer at Shopify, Dec 2024 - Present · 1 yr 9 mos, India, Empowering merchants with AI based companions, their own Sidekick.

Senior Data and Applied Scientist at Microsoft, Sep 2023 - Dec 2024 · 1 yr 4 mos, Vancouver, British Columbia, Canada, I contributed to a horizontal initiative to integrate unified embeddings for cross-product personalization, creating pipelines that generate 1M to 50M user profiles daily. I integrated LLM and RAG evaluation schemes, accelerating iteration speed and impacting over 400M users across three products. As Tech Lead, I also diversified and enhanced image recommendations using Bing and Edge signals, using LLM-based retrieval algorithms to boost retention and engagement by 7-9%, and reduced in-session and cross-session duplicates, cutting Total Time on Server by around 15%. I also integrated multi-modal feature embeddings, increasing Engaged Sessions per User by 2.21% and Daily Active Users by 3.03%, and developed scalable, modular code, reducing flighting time from 3-4 days to 0.5 day.

Data and Applied Scientist at Microsoft, Jul 2021 - Sep 2023 · 2 yrs 3 mos, Vancouver, British Columbia, Canada, Working in Bing Multimedia team at Microsoft.

Associate Researcher at Huawei, Feb 2020 - Jul 2021 · 1 yr 6 mos, Toronto, Canada Area, Tools Used: C++, Python, Tensorflow, Tensorflow Lite , Android JNI, HiAi Platform, ONNX • Prototyped face detection and object detection for android systems. • Mentored four co-op intern students working on different ML inference projects • Contributor to drowsiness detection system, featured as a major product by Huawei CBG CEO • Consulted on multiple projects for architectural design and embedded systems related optimization.

Machine Learning Research Intern at Phenomic AI, May 2019 - Dec 2019 · 8 mos, Toronto, Canada Area, Tools Used: Python, CellProfiler, AWS services The project aims at applying machine learning in the area of drug discovery. Over a thousand microscopy images get generated every week and eye-based analysis is tedious. Hence, to make the process efficient, the segmentation and the feature extraction process is automated using various AWS services and latest cellular segmentation algorithm. After the individual cellular area are identified and their features are extracted, it is followed by analysis to classify the cells in images into different cell types using suitable supervised or semi-supervised learning techniques. The analysis also involves validation and quantification of the observation made by biologists to understand the effect of the environment on which the cells are put to. --------------------------------------------------------------------------------------------------------- BACKGROUND Cell-cell interactions have been found to have a significant impact on drug efficacy. However, various experimental conditions are tested and manual analysis quickly became a bottleneck. My main aim is to analyze the microscopy images of multi-cell cultures and extract intensity, shape, and texture-based features using an automation pipeline. These features are then put for analysis to validate the observation made by the biologists or understand the effects of the environment on the cells in which they are put.

Graduate Teaching Assistant at University of Toronto, Sep 2018 - Apr 2019 · 8 mos, Toronto, Canada Area, For the course: Computer Networking Systems (Fall Term), Intro to Programming (Winter Term)

Research Engineer at Samsung, Jun 2017 - Aug 2018 · 1 y 2 mo, Bengaluru

Education

Master of Science in Applied Computing, Computer Science, University of Toronto

Bachelor’s Degree, Computer Science and Engineering, Indian Institute of Technology, Guwahati

Public links

github: https://github.com/ranka47