Tanmay Gupta

Full-Stack ML Engineer

I build, train, and deploy ML models, design and build web apps, and find ways to market / sell it.

My Story

I was born in a military family, which meant moving every 2 years, I lived in over 20 cities / naval bases across India, making friends, picking up local tastes, and moving on.

At 12, I entered boarding school in the Himalayas, embarking on a six-year journey of self-discovery and deep friendships, where mountain treks became my escape, friends became family, and the school became my world.

I moved to New York City for college, starting in Finance at NYU before an internship rekindled an interest in Mathematics and led me to machine learning. I worked all 4 years of college in various part-time jobs - starting as a service-desk agent at NYU IT and ending as a Data Scientist in NYU nursing school.

I then took on full-time roles at IBM in their Chief Analytics Office developing ML models for IBM Cloud. Following my passion for e-commerce while building my mother's online store Incense Art Studio, I joined Shopify as a Data Scientist, working on ML models for Inventory distribution.

Following visa expiry in 2023, I returned to India after 7 years. Made the most of it, visiting 15+ cities, traveling to the edges of the country, and reconnecting with family and friends and building Quarkle with my high school friend.

Experience

  1. Oct 2023 - Present

    Co-Founder and CTO ⚡️ Quarkle Inc.

    🔗: https://quarkle.ai/

    Quarkle is an AI platform that provides developmental feedback to self-published authors. Scaled to 3500+ users in 6 months with $15K ARR. Python based FastAPI backend and AI service deployed on AWS and Next.js frontend.

    • REACT
    • NEXT.JS
    • NODE
    • AWS
    • JAVASCRIPT
    • TAILWIND
    • PYTHON
    • DOCKER
  2. Jan 2022 - June 2023

    Data Scientist ⚡️ Shopify Logistics AI Team

    Part of the Inventory Planning and AI team at Shopify. Built and deployed hierarchical probabilistic inventory allocation models for 300K products across 50+ warehouses. Worked on CEO's special projects experimenting with H3 geospatial indexing and transformer forecasting models.

    • PYTHON
    • SQL
    • SPARK
    • PYTORCH
    • GOOGLE CLOUD
  3. Aug 2020 - Dec 2021

    Data Scientist ⚡️ IBM Chief Analytics Office

    Researched and engineered monthly usage forecasting model for IBM Cloud customers for 500+ Customer Success Managers. Implemented and customized research paper by Hyndman et al. on Feature-based Forecast Model Averaging for Time Series Ensembling using tree-based ensemble methods.

    • PYTHON
    • SQL
    • IBM CLOUD
    • DOCKER

👉 Connect with me on LinkedIn

Projects

  1. July 2024 - July 2024

    Keen-Eyes

    Keen-Eyes is a cli tool to give feedback on your git commits and PRs using AI.
    • BASH
    • LLM
  2. Feb 2024 - May 2024

    Jay.io

    Jay started as an internal tool to function as a headless email copilot within slack. It has now evolved into an AI-coworker which can write emails, schedule meetings, manage your calendar, and perform research.
    • CREWAI
    • AGENTS
    • SLACK
    • AUTOMATION
  3. June 2024 - June 2024

    Formie

    Formie is a google-chrome extension that allows users to fill out forms faster by tracking and recommending answers based on user's previous responses.
    • CHROME EXTENSION
    • WEBLLM
  4. Feb 2024 - Feb 2024

    PriomptiPy

    PriomptiPy is a Python library that integrates priority-based context management into prompts for LLMs to ensure that the LLM is using the most relevant context given limited context.
    • LLM OPS
    • OPEN SOURCE
  5. June 2023 - Aug 2023

    HuberX

    HuberX is a GPT based AI QA tool that answers questions based on Dr. Huberman's podcast episodes. It is built using Hugging Face's transformers library and deployed on Vercel.
    • RAG
    • CHATBOT

👉 View more on GitHub