Ishan Vepa

engineer intern @ adobe | cs @ gatech

CS student at Georgia Tech studying Human-Computer Interaction and Artificial Intelligence. When I'm not programming, you may find me transcribing jazz records, trying new shaken espressos, and watching Premier League soccer (GGMU).

Experience

Adobe

Software Engineer Intern

May 2026 – Aug 2026

Mitigated bot abuse across Adobe’s commerce services with ML-based risk detection, flagging XXK+ abusive transactions daily and preventing $XXM+ in annualized fraud and compute losses
Engineered LightGBM fraud detection model scoring 200k+ transactions/day with 86% precision and 91% recall for identifying fraud, reducing false positives by 27%
Built an automated model retraining pipeline via Databricks cron job over XM+ transactions, engineering proprietary features, tuning with Optuna, and serving in real-time via internal inference service
Developed ARP Policy SDK: an agentic fraud-policy auditing service identifying duplicate/overlapping rules via Jaccard similarity, simulating impact over Databricks data, and drafting rule-engines from natural language
Skills: Python, LightGBM, Optuna, Scikit-Learn, Spark, Databricks, Jupyter Notebook, OpenAI API, SQL

SitScape

Software Engineer Intern

Jun 2025 – Aug 2025

Engineered a Generative UI system with a multi-agent AI ReAct service using LangGraph/LangChain in Python that automates front-end UI and event handling, reducing developer build time by 75%.
Shipped a no-code RAG API, enabling 10+ Fortune 500 clients to deploy LLM-powered chat in production.
Authored comprehensive documentation on both services for use by a team of 30+ developers.
Skills: Python, LangGraph, LangChain, Flask, JavaScript, SQL, OpenAI API.

Datamaxx Group

Software Engineer Intern

May 2024 – Aug 2024

Reduced user identity verification time by 67% (1 minute to 20 seconds) by developing an automated ID parsing and user face verification service in ASP.NET with AWS Textract and Rekognition API.
Enhanced user security screening efficiency for 100+ vendors by implementing a QR code visitor check-in system.
Skills: C#, ASP.NET MVC, jQuery, HTML, CSS, SQL, AWS API.

Education

Georgia Institute of Technology

M.S. Computer Science: Human-Computer Interaction

Jan 2026 – Present

Georgia Institute of Technology

B.S. Computer Science

Aug 2022 – Dec 2025

Research

Georgia Tech Sonification Lab & Design Studio

Graduate Researcher

Conducting HCI research on accessible big data systems.

Aug 2024 – Present

Publications

Ishan Vepa, Jessica Herring, Isabella Martincic, and Bruce N. Walker. 2026. Inclusive Climate Communications: Accessible Hurricane Tracking Maps for the Web. Web4All 2026: 23rd International Web for All Conference (W4A '26). April 13–14, 2026, Dubai, UAE. ACM, New York, NY, USA. 2 pages. https://doi.org/10.1145/3800424.3800436