Loading portfolio experience
Loading portfolio experience
Business Operations, Education · AI Agent, Web App, MVP
An agentic AI application that collects news and scientific articles from configured feeds, organizes them by keyword and topic, removes duplicates, and provides source-grounded summaries.
Researchers needed to monitor news publications and scientific feeds for selected keywords, but relevant articles were spread across numerous sources and often repeated the same development.
I developed an MVP using the OpenAI SDK and a set of specialized agents. Collection agents retrieve articles from approved feeds, processing agents extract metadata and remove duplicates, and analysis agents classify topics and create evidence-linked summaries. Users can search, filter, sort, save, and review articles through a lightweight interface. The architecture, prompts, agent responsibilities, and development process were documented as an educational case study.