Sathvik Chinta, Developer in Pittsburgh, PA, United States
Sathvik is available for hire
Hire Sathvik

Sathvik Chinta

Verified Expert  in Engineering

Machine Learning Engineer and Software Developer

Location
Pittsburgh, PA, United States
Toptal Member Since
September 5, 2023

Sathvik is a master's student proficient in data science and software engineering. He specializes in machine learning and artificial intelligence but has extensive experience in cloud systems and full-stack and back-end development. Sathvik can deliver complex solutions using Java, Python, C++, Kotlin, and Go.

Portfolio

Coupang
Artificial Intelligence (AI), Machine Learning, Data Scientist, PyTorch...
Coupang
AWS IoT, Python 3, Amazon EC2, Amazon S3 (AWS S3), OneDrive, TensorFlow...
Coupang
SQL, Apache Hive, Zeppelin, Amazon EC2, Amazon S3 (AWS S3), PyTorch...

Experience

Availability

Part-time

Preferred Environment

Visual Studio Code (VS Code), Linux, Python 3, MacOS, Windows

The most amazing...

...thing I've developed is an internal chatbot network using custom LLMs that can communicate with each other to answer the user's questions.

Work Experience

Data Scientist

2023 - 2023
Coupang
  • Started and spearheaded a new project as a full-time data scientist on the Rocket Growth team to advance the platform to the next level by utilizing custom-built large language models (LLMs).
  • Theorized, ideated, and deployed new ways for LLMs to interact with one another to deliver answers quickly and securely.
  • Created models with PyTorch to identify bundles in the eCommerce marketplace.
  • Used cloud technologies such as Amazon S3, Amazon SageMaker, and Amazon EC2 actively.
  • Developed cutting-edge models using open-source repositories like Hugging Face and GitHub.
  • Suggested, prototyped, and documented multiple high-value use cases for new technologies for the team.
  • Used multiple Python libraries, including pandas, Polars, and NumPy.
Technologies: Artificial Intelligence (AI), Machine Learning, Data Scientist, PyTorch, TensorFlow

Data Science Intern

2022 - 2022
Coupang
  • Spearheaded a cross-team project using computer vision algorithms to identify bad actors in the eCommerce marketplace.
  • Identified brands by leveraging PyTorch and popular machine learning models such as YOLO.
  • Used cloud resources like Amazon EC2, Amazon S3, and Microsoft OneDrive.
Technologies: AWS IoT, Python 3, Amazon EC2, Amazon S3 (AWS S3), OneDrive, TensorFlow, TensorBoard

Data Science Intern

2021 - 2021
Coupang
  • Collaborated with the visual intelligence team on a project to improve product categorization. Learned how to read, contextualize, and implement topics from emerging research papers.
  • Used new frameworks and techniques, including Amazon EC2, Amazon S3, PyTorch, SQL, Hive, Zeppelin, Vim, and Linux command line.
  • Handled large-scale datasets with billions of rows and used SQL to effectively combine tables and create training, validation, and testing datasets.
  • Implemented an end-to-end ML pipeline for our models in PyTorch, utilizing both text and image inputs to make classifications.
  • Learned how to create neat, modular, and easily reproducible code for ML training. The code allowed for a plethora of configurations and made it easy to mix and match according to the end user's wishes.
  • Implemented groundbreaking techniques, such as automatic mixed precision, distributed data parallel, and decision-level fusion techniques.
  • Hosted and trained a model on AWS, extensively using Amazon EC2 and Amazon S3. Collaborated with another intern to divide and conquer model production.
Technologies: SQL, Apache Hive, Zeppelin, Amazon EC2, Amazon S3 (AWS S3), PyTorch, Vim Text Editor, Linux

Software Engineering Intern

2021 - 2021
Amazon.com
  • Collaborated with the Amazon Lookout for Metrics team and learned different AWS products in a very short time.
  • Automated user entries in Lookout for Metrics by taking a CSV and JSON input file and inferring all fields. It would then be automatically filled in on the website and returned to the user for confirmation.
  • Created an API using Amazon's internal builder tools to interact with their service. Worked in a large team with multiple members to achieve the best results.
Technologies: CSV, JSON, AWS IoT, Kotlin

Software Engineering Intern

2020 - 2020
Coupang
  • Built API tests in the Java source code using Spring and JUnit. The tests interacted with our database and ensured everything was made correctly and functioned as intended.
  • Learned new frameworks and applications such as JUnit, Git, Spring, Google Puppeteer, and TensorFlow, as well as the DevOps lifecycle and the Agile development approach in the industry environment.
  • Provided an entire DevOps pipeline for the team utilizing Jenkins and other continuous integration tools.
  • Created my build jobs that hosted end-to-end tests and worked with Groovy pipeline scripts to trigger and provide feedback automatically.
  • Used Puppeteer, automated web manipulation software, to create an end-to-end test for our service and the Jest framework to test and assert. All the tests were hosted on a custom Jenkins build.
  • Suggested, learned, and created TensorFlow models for ML purposes to elevate the platform to the next level. Used neural networks and boosted trees for different models.
  • Created tools with React to interact with multiple APIs and display information from our database in a clean and organized way.
Technologies: DevOps, Agile DevOps, JUnit, Java, Git, Spring, Puppeteer, Back-end, Full-stack, TensorBoard, TensorFlow, Python 3, Jenkins, React

Seattle PD Crime Analysis

The aim was to create a data science project to analyze the Seattle Police Department (PD) crime reports. My team cleaned, analyzed, and contextualized the data.

As part of the team, I built machine learning models to predict crime counts for different crime categories using TensorFlow to a high degree of accuracy. I also used dataset partitioning methods such as moving windows. In order to compare and contrast different methods and find ideal models for all crime categories, I utilized linear models, deep neural networks, convolutional neural networks, and recurrent neural networks with long short-term memory blocks.

The project resulted in winning 3rd place for the best machine learning model.

Mingle

https://github.com/NSC508/Hack-20/tree/master
A web app for connecting people during quarantine. It matched people with similar interests and classes and supported class study groups and messaging.

The front end was built with React and the back end with Python. I developed the back end from the ground up and implemented Firebase in the web app. The end product used Google Authenticator for login and Firestore for the database. I wrote Python code from scratch, allowing interaction with the database to add, query, and logically identify people with similar interests. For hosting my web server, I used Flask and created an API to act as an intermediary between my Python code in the back end and React code in the front.
2019 - 2023

Bachelor's Degree in Applied and Computational Mathematical Sciences

University of Washington - Seattle, WA, USA

Libraries/APIs

TensorFlow, PyTorch, OneDrive, Puppeteer, React

Tools

TensorBoard, Vim Text Editor, Git, Jenkins

Languages

Python 3, Java, SQL, C++, Kotlin

Platforms

Visual Studio Code (VS Code), Linux, MacOS, Windows, Docker, AWS IoT, Amazon EC2, Zeppelin, Firebase

Frameworks

JUnit, Spring, Flask

Storage

Amazon S3 (AWS S3), JSON, Apache Hive, Google Cloud

Paradigms

DevOps

Other

Data Structures, Machine Learning, Data Scientist, Software Development, Linear Algebra, Algorithms, Supercomputers, Artificial Intelligence (AI), CSV, Agile DevOps, Back-end, Full-stack

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring