Education and Experience

I'm a results-driven ML researcher with 3+ years of experience driving impactful business outcomes through machine learning. At Goldman Sachs, I tackled critical challenges by building and deploying cutting-edge models for credit decisions and market analysis. My work led to tangible improvements, including a 6% decrease in delinquency rates and actionable insights for policy adjustments. I'm proficient in building and deploying models on platforms like Provenir, and my strong academic background and ongoing research at UC Santa Cruz fuel my passion for innovation.

Sumary

Rithesh Kumar

Highly motivated researcher with experience in developing and deploying machine learning models for financial services and social media analysis.

  • Santa Cruz, CA
  • (831) 266-7201
  • rkumar45@ucsc.edu

Education

Master of Science in Computer Science

2023 - Present

University of California Santa Cruz, Calofornia, USA

GPA: 4.0/4.0

Courses: NeuroSymbolic AI, Advanced Topics in NLP: Generative AI, Artificial Intelligence

  • Teaching Assistant for Computer Architecture (Jan 2024 - Mar 2024)
  • Teaching Assistant for Ethics and Algorithms (Apr 2024 - Present)

Bachelor of Technology in Computer Science

2016 - 2020

National Institute of Technology Karnataka, Surathkal, INDIA

GPA: 9.39/10.0

Courses: Computer Vision, Data Mining and Visualisation, Graph and Number Theory, Probability and Statistics

Professional Experience

Artificial Intelligence Intern

2024

Accretional, San Francisco, CA

  • Contributed to the early-stage development of Brilliant, a cutting-edge AI tool designed to simplify API writing and deployment.
  • Implemented key features enabling the product to seamlessly generate and deploy APIs to Google Cloud Functions and AWS Lambda, enhancing operational efficiency.
  • Fine-tuned and engineered prompts for large language models (LLMs), optimizing their performance in generating accurate and efficient API implementations.
  • Currently working on a custom search engine focused on semantic search across existing APIs, enhancing the model's capabilities through Retrieval Augmented Generation and reducing hallucination in model outputs.

Decision Scientist

2020 - 2023

Goldman Sachs, Bangalore, INDIA

  • Developed 2 models based on SVM and Knowledge-Tree graphs for customer delinquency and risk scoring for personal loans and further assisted in its deployment. The addition of the new models resulted in a 6% decrease in the overall delinquency rate and an increase in customer satisfaction by 8%
  • Built 2 regression models to assist in customer acquisition risk scoring and personalize the loan offers for each customer. Deployment of these models led to an increase in the loans provided by 5%
  • Spearheaded the research on changing market conditions during the COVID period by leveraging the tools of LSTM-based time series analysis over the years. The findings of the study were then used to refine the existing policies, leading to an increase in new customers by 10%
  • Supervised a team in building a comprehensive end-to-end pipeline for the credit decision and loan offer process on Provenir, leading to a decrease in the latency of loan processing by 15%, an increase in load handling by 13% and reducing the cost of maintenance of the pipeline by 20% compared to the previous pipeline deployed on a legacy platform, FICO.

Machine Learning Research Intern

2019

Sprinklr India, Gurgaon, INDIA

  • Developed and deployed an LSTM model that reduced spam classification errors by 10%, significantly improving data quality for product review analysis.
  • Conducted research on model compression for LSTMs, proposing a teacher-student model that achieved a 60% size reduction with minimal accuracy loss. This project demonstrates your problem-solving skills and understanding of model efficiency, which is crucial for large-scale deployments.
  • Analyzed and modeled social media data (Twitter) using LSTMs, showcasing your ability to work with complex datasets relevant to Amplitude's core business.