Work:
- RAG Citation Generator
- Design and implement novel Large Language Model Citation Generator using multi-Gaussian clustering algorithms and dynamic content similarity techniques, yielding 80% more relevant document references for production-grade RAG model.
- SQL Retrieval Agent
- Develop proof-of-concept few-shot LLM prompting framework to translate natural language queries into executable SQL, integrating regex-based syntax validation, enhancing table-based document retrieval and question answering.
- SAS Code Translation Agent
- Augment state-of-the-art AI Code Assistant tools to create a novel LLM-powered SAS to python IT chain management code translation tool. Build and integrate Retrieval and Action (ReAct) agent for SAS code analysis and equivalent python architecture generation to guide agentic translation pipeline.
- Upskilling the Workplace with AI Skills
- Create week long lecture series and corresponding interactive educational material for both company internship program and company-wide audience aimed at providing general overview of GenAI and focusing on how to use AI to optimize one’s workday. Collaborates with educational team and delivers lecture series, focusing on understanding LLMs at a conceptual level and mastering techniques such as prompt engineering, model tuning, workflow building, and AI-assisted coding.
- Smart Scrum Master - 2025 Rocket.Build Hackathon Finalist
- Create an LLM-powered application that interfaces with daily Atlassian developer tools to optimize agile scrum rituals aimed to organize, augment, and track team workload, progress, and velocity.
- Aspera API Test Integration
- Scripted in Ruby to automate API endpoint testing for entire IBM Aspera on Cloud Workflows App designed to automate workflows of global content transfer and exchange across on-premises and multi-cloud environments.
- Developed and documented comprehensive automated regression framework with 780+ API tests covering all endpoints.
- Identified 15+ critical issues and bugs on live production server, communicated with developers to determine ideal expected behavior, and raised issues in project management tools to expertly report relevant failure details.
- Executed full development cycle to outline and implement API Endpoint Test Controller for collaborative QA repository.
- MultiValue Experience Unit Test Coverage
Research:
-
publication: https://openlib.tugraz.at/download.php?id=686289f5ddb80&location=browse
- Pirat Homebase
- Robotics Using Bayesian Inference Lab
- Chiba Lab
- Pirat RL Regulation Optimization
- Robotics Using Bayesian Inference Lab
- Chiba Lab
- AI Classroom
- Center for Human Development
- Cognitive Science Department
- ABCD - Adolescent Brain Cognitive Development Study, UC San Diego Coordinating Center