As a data science intern at CCC Intelligent Solutions, I engineered prompts for visual language models to enable faster and accurate damage summarization for insurance claims. I also deployed a pipeline to remove PII in both visual and textual data, enhancing the feature store for inbound subrogation models. Additionally, I created a package of metrics to evaluate the quality of generated text with ground truths.I graduated with my Master's degree from the University of Southern California, with a focus on machine learning and deep learning. I have a bachelor's degree in computer engineering from Thadomal Shahani Engineering College, where I worked on small-scale visualizations and deep learning projects. I have gained multiple certifications in deep learning for computer vision, artificial intelligence, and neural networks from reputable organizations. I am proficient in RNN, Angular, and Material-UI, among other skills.My goal is to further accustom myself with ML algorithms and deep learning frameworks, and to apply them to real-world problems. I am looking for full-time roles in ML/DS/SWE, where I can leverage my skills and experience to contribute to the team and the organization. I am passionate about building a career around these fields.
-
Software EngineerPalletLos Angeles, Ca, Us -
Lead Software DeveloperYour.MovLos Angeles, Ca, Us -
Founding Software EngineerStealth Startup Mar 2024 - PresentLos Angeles, California, United States -
Data Science InternCcc Intelligent Solutions Aug 2023 - Dec 2023United States• Engineered prompts for IDEFICS and Cheetah (Llama Based) visual language models for the task of Visual-Question Answering for damage summarization, helping clients understand claim information in seconds instead of minutes.• Deployed into production a pipeline to remove PII in both visual and textual data with help of AWS Step Functions, Lambdas, CloudFormation, EventBridge., scheduled to be nightly batched. The pipeline enhances the feature store being used by the Data Science department to create inbound subrogation models and is expected to process 100000 demands per week.• Created a package of metrics to evaluate generated text with ground truths. The package provides similarity-based metrics like BERT embeddings as well as metrics relying on distance measures like fuzzy edit distance and will be used to evaluate outputs of all generative language models being developed by the company.• Worked on a machine learning model pipeline using CLIPSeg for evidence verification in subrogation claims to help increase success rates in subrogation disputes.• Wrote unit tests and regression tests with benchmarks using pytest and pytest_benchmark to ensure smooth functionality of code. -
Robotics Instructor And Program CoordinatorUsc Information Sciences Institute Jun 2023 - Jul 2023Los Angeles, California, United StatesRobotics Instructor and Program Coordinator for the USC ISI x Amazon Stimulating Stem Summer Program -
Co EditorFresh Off The Field Jan 2020 - Dec 2021Mumbai, Maharashtra, India -
Machine Learning InternCloud Counselage Pvt. Ltd. Mar 2020 - Jul 2020Mumbai, MaharashtraDeveloped an Event Recommendation System to extract the domain and event-type from an event headline and then recommend the event to employees using concepts like tf-idf and Support Vector Classifiers. I was also responsible with creating the dataset for the training as well as submitting timely reports
Shivam Kotak Education Details
Frequently Asked Questions about Shivam Kotak
What company does Shivam Kotak work for?
Shivam Kotak works for Pallet
What is Shivam Kotak's role at the current company?
Shivam Kotak's current role is Software Engineer.
What schools did Shivam Kotak attend?
Shivam Kotak attended University Of Southern California, Thadomal Shahani Engineering College.
Not the Shivam Kotak you were looking for?
-
Shivam Kotak
Ahmedabad -
Shivam Kotak
International Affairs & Economics Student At The George Washington UniversityWashington, Dc -
-
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial