AIP Pipeline Builder: April 2023 - April 2024Implemented a DSL (domain-specific language) for application context, enabling LLM model communication and supporting code generation and descriptions.Developed a universal testing framework for LLM features, enhancing iterative improvements across various test sets(SPIDER) and metrics (pipeline compilation, anti-hallucination heuristics).Foundry Pipeline Builder: Sept 2022 - April 2023Contributed to the core Pipeline Builder, a no-code/low-code tool for data transformations and real-time data preview.Designed a versioned internalization framework for backend service translations.Successfully implemented input sampling, significantly reducing preview times.

Teaching Computer Programming and Automata Theory.

- Worked on the development of Foundry's datasets' build infrastructure, implementing diverse failure strategies - Designed and deployed multiple APIs within a micro-services architecture

- Developed a commodities product recommender for clients by using product metadata and anonymized user’s data- Implemented the recommender in Keras using matrix factorization and content-based filtering- Monitored user behavior using internal UI frameworks and Grafana

Developed a tool to clean the dataset. Separated and parallelized different jobs in the training pipeline (evaluate jobs, log parsing, model backups). Worked on post processing object detection neural network noisy outputs(clustering, KDE).

Developed an Asp.Net web application using MVC framework, Entity Framework, CSS and JQuery.

Developed an Android-based application using ASP.NET Core.