Sarthak Gupta

Co-Founder

Charlotte, North Carolina, United States4 yrs 1 mo experience
AI EnabledAI ML Practitioner

Key Highlights

  • Optimized ML inference time from 7 hours to 36 minutes.
  • Led a $50k pilot project for a $1B+ pharma company.
  • Founded a social app aiming to redefine user engagement.
Stackforce AI infers this person is a SaaS and Healthcare-focused Software Engineer with strong ML and UX capabilities.

Contact

Skills

Core Skills

Machine LearningData EngineeringSoftware EngineeringBusiness OwnershipArtificial Intelligence (ai)User Experience (ux)

Other Skills

Python (Programming Language)Google Cloud Platform (GCP)DockerApache BeamML InfrastructureModel PlatformNeo4jReact.jsJavaREST APIsProblem SolvingCommunicationAmazon Web Services (AWS)MySQLNode.js

About

I’m a product-minded SWE/ML engineer and founder who ships systems that measurably move revenue and efficiency. I don’t wait for problems to land on my desk; I look for gaps in systems early and turn them into opportunities to improve performance and reliability. By 19, I’ve led an ML platform project at Intuit Credit Karma, a $50k enterprise SaaS pilot for a $1B+ pharma company, and built full-stack products from social apps to edtech and health-tech. My sweet spot is the intersection of software engineering, ML, data infrastructure, and UX, translating complex pipelines into clean, scalable solutions that directly tie to business outcomes. I’m looking for teams where I can own meaningful pieces of the stack, move fast without breaking trust, and compound impact across teams.

Experience

4 yrs 1 mo
Total Experience
9 mos
Average Tenure
--
Current Experience

Credit karma

Software Engineer - Machine Learning Infrastructure/Platform

May 2025Aug 2025 · 3 mos · United States

  • Our ML Platform drives 60 Billion/day predictions across our 130M users.
  • 1/ Optimized ML batch inference pipeline on Apache Beam / GCP Dataflow, cutting scoring latency 7 hrs → 36 mins 92% faster and reducing cost $600 → $89 (85% savings), unlocking $400k+ annual savings.
  • 2/ Implemented batch scoring service in Python (PMML runtime, schema validation, feature normalization) packaged as Docker image and deployed via Dataflow Flex Templates. Integrated with CircleCI and merged to master for team-wide reuse.
  • 3/ Scaled batch inference to 60M records/day across 40 workers via benchmarking + tuning (batching strategies, JVM heap sizing, downstream filtering) to maximize throughput while controlling spend.
  • 4/ Enabled 10x more model onboarding capacity for Data Science teams (e.g., AlgoFacts, TaxML) without proportional increases in compute load or spend.
  • 5/ Applied systems-level thinking to production ML, balancing latency, throughput, and cost tradeoffs at 9-figure scale.
  • Edit: A similar optimization strategy is now appearing in Google Cloud Platform for ML inference - I helped Intuit Credit Karma get there first, in 11 weeks
  • Source: https://cloud.google.com/bigquery/docs/inference-overview
  • EDIT: We followed OpenAI's SWE Engineer Project Steps:
  • a.) Defining a clear hypothesis (“we think X will reduce latency by Y under Z”)
  • b.) Instrumenting existing production systems, gathering metrics and detailed analysis to validate the hypothesis
  • c.) Building or modifying a real system (prototype or production-quality improvements when appropriate)
  • d.) Running experiments/benchmarks and analyzing results
  • e.) Communicating tradeoffs and recommendations clearly
  • f.) Publishing the research work in technical journals and conferences
  • Credit to my talented engineering team and directors: Pallav Anand, Nicholas Pataki, Debasish Das, Hardik Udeshi, Raj Katakam, Dharam Pulakunta, Madelaine Daianu
Python (Programming Language)Google Cloud Platform (GCP)Data EngineeringDockerApache BeamMachine Learning+1

Intuit

Machine Learning Engineer

May 2025Aug 2025 · 3 mos

  • Parent company of Credit Karma.
Python (Programming Language)ML InfrastructureModel PlatformMachine Learning

Docnexus (techstars '22)

Software Engineer, Enterprise Software

May 2024Aug 2024 · 3 mos · Seattle, WA

  • At 16 y/o I brought in $50,000 in revenue and contributed to a $1M+ pipeline. Here's how:
  • Developed an ML-powered data ingestion pipeline using spaCy and FastAPI to automate entity extraction from large-scale medical publications datasets (e.g. drug names, diseases, author).
  • Spearheaded the development and deployment of 5,000+ lines of code for a $50k pilot with a $1B+ pharmaceutical company, building an AI-powered research platform that accelerates drug discovery for Myeloma cancer patients by surfacing insights from vast medical datasets.
  • Contributed to a predictive model for user engagement using Scikit-learn, experimenting with logistic regression and random forests to analyze clickstream data.
  • Implemented a CI/CD pipeline using GitHub Actions for a React webapp and API services; automating builds, tests, and deployments to AWS, reducing manual deployment time.
Python (Programming Language)Neo4jReact.jsJavaREST APIsProblem Solving+16

1700ventures.com

Software Engineer, B2B SaaS

Jan 2024May 2024 · 4 mos · North Carolina, United States

  • (b2b SaaS) An end-to-end platform for student success, acting as the OS (Operating System) for Student Lifecycle Management by automating routine administrative tasks, reducing errors, and freeing up time for more impactful, personalized student engagement.
  • Developed an AI-based scheduling tool to optimize course scheduling and student support, by integrating existing Python-based ML models for personalized student support recommendations.
  • Implemented a rule-based validation system using Python and Pandas, reducing manual record correction time for administrative staff.
  • Created ReactJS frontends and Node.js backend, developing interactive webpages for school district dashboards.
  • Worked with AWS Lambda to optimize API calls for a cloud-based SaaS platform, reducing load time by ∼15%.
  • Implemented AES-256 encryption for secure data transmission, ensuring end-to-end security of sensitive database fields, by researching security best practices and writing unit tests for data encryption functions.
  • The OS for School District Administrative Auditing
Problem SolvingBusiness OwnershipEnterprise SoftwareSoftware Architecture & DesignReact.jsArtificial Intelligence (AI)+3

Spacesnap

Founder, SWE + UI/UX + Marketing

Jan 2023Feb 2024 · 1 yr 1 mo

  • I built a social app that could dethrone Instagram and Snapchat. Here's what happened:
  • The vision was simple but revolutionary:
  • Social media that puts you IN THE MOMENT
  • Your posts appear WHERE you create them
  • Memories that STAY instead of disappearing into the endless feed
  • What I built from scratch:
  • Full-fledged iPhone app deployed to the App Store
  • Real-time posting functionality that real users actually used
  • Location-based memory pinning (your posts stay where you create them)
  • Voice conversations as the primary communication method
  • Complete backend infrastructure handling thousands of requests
  • Tech stack I mastered along the way:
  • React Native for cross-platform development
  • PostgreSQL database architecture
  • Cloudflare for scalable deployment
  • Twilio integration for SMS notifications
  • Apple TestFlight for beta distribution
  • Traction that proved it wasn't just another app idea:
  • Grew dedicated email and SMS subscriber list
  • Secured beta testers through Apple TestFlight
  • Conducted user interviews that shaped the roadmap
  • Instagram captures moments.
  • Snapchat makes moments disappear.
  • Spacesnap makes moments matter.
  • Temporarily pausing marketing to build a health-tech app for children with autism, integrated with invisible AI. It’s a novel approach that puts kids in the driver’s seat of their social development.
User Experience (UX)React NativePostgreSQLApple TestFlightGo-to-Market StrategyTwilio+3

Packed

Software Engineer, Order Delivery Software

Sep 2022Jan 2023 · 4 mos · San Diego, California, United States

  • The Doordash hyper-specific for college students!
  • Ensured smooth app functioning under heavy usage by resolving system validation tests, resulting in improved user experience and increased customer satisfaction.
  • Designed, planned, and integrated cloud computing and virtualization systems using code and modern cloud-native deployment techniques, resulting in improved scalability and performance.
  • Developed the app using Model-View-Controller (MVC) design patterns, Software Requirements Specification (SRS), and Agile software development methodologies, resulting in faster development times and higher-quality outcomes.
  • Formulated CRUD operations through cloud services and a cloud backend developed in Node.js, resulting in improved database management and increased data accuracy.
  • Utilized SQL and NoSQL databases and strong use of JSON objects within the application and cloud, improving data handling and management.
  • Led integration of REST APIs into the React Native app and used modern async await methods for optimal performance, resulting in faster app response times and better user experience.
  • Utilized React Navigation for app-wide navigation and Redux for app-wide state, improving app organization and management.
  • Conducted detailed code reviews of pull requests, ensuring clean code, optimization, and no bugs, resulting in improved code quality and better team collaboration.
  • Developed the application feature by feature in strict time frames and ensured high quality while working on the respective feature, resulting in faster development times and better outcomes in a fast-paced agile environment.
JavaReact NativeProblem SolvingCommunicationDatabasesGitHub+7

Oxidezz

"Full-Stack" Entrepreneur - Take 1 of Business 🎬

Dec 2020Aug 2021 · 8 mos · United States

  • Billionaires hit mistakes and think how to do better in their next venture. Worked on building a full-on clothing brand for teens working on everything from start to finish. Read ahead...
  • Successfully launched the store, but learned an important lesson that sticks with me to this date: even though my marketing was really good and the target market's feedback was positive, I should have first launched a round of pre-orders to assess if people were truly willing to purchase the clothing.
  • I learned:
  • Building an ecommerce store along with the nitty-gritty features that Shopify offers.
  • Using payment processing.
  • Creating realistic product images without having the physical clothing - intricate photo/video editing.
  • First time creating and setting up a Facebook Ads Manager Account.
  • Convening with potential teens interested in the brand for feedback and iterating over design ideas.
  • Building branded SMS automations for notifying people of their order.
  • Communicating for hours with Print-on-Demand (POD) suppliers to understand their operations and output, pricing structure, taking note of helpful suggestions by their CS team, clarifying questions about their platform.
  • How to quickly lose sleep
  • Instead of hustling to work on one-task-after-another, scope out and look at the bigger picture to understand if what I’m working on a day-to-day basis is actually working towards achieving the bigger goal.
  • All of this, self-taught.
  • Important lessons learned; now building the next big thing with these lessons in mind.
CommunicationBranding & IdentityHow to lose sleepE-CommerceSupplier SourcingBusiness Ownership

N/a

Programming Skill-Set Enhancement

Feb 2020May 2022 · 2 yrs 3 mos · United States

  • Overall, I touched on all aspects surrounding software development: Python, Java, AI/ML, Automation...
  • ● Dedicated 100+ hours enhancing Java programming skills through real world projects such as a school management system. Additionally worked with Android Studio to create Android native apps in Java. Wrote 1,000+ lines of code.
  • ● Improved skill set in Python programming by learning Selenium - a web-based automation tool which is beneficial in website testing.
  • ● Effectively learned relevant technologies for artificial-Intelligence skills through OpenCV, for face detection, car detection, and object detection.
JavaTest AutomationGitHubTeam LeadershipAndroid StudioPython (Programming Language)+3

Education

University of North Carolina at Charlotte

Bachelor of Science - BS

Harvard University

CS50 — Computer Science

Apr 2020Jul 2020

Cox Mill High School

Computer Science & Leadership

Udemy

Computer Science — Java

Stackforce found 100+ more professionals with Machine Learning & Data Engineering

Explore similar profiles based on matching skills and experience