The final unit was the most personally rewarding. My individual development project focused on building a research agent that could assist with discovering academic papers. I used tools like Streamlit, Beautiful Soup, and SQLite to build the interface and backend, and incorporated SerpApi for search logic.
This project was a culmination of everything I’d learned, from designing agent architecture to handling uncertainty, and adaptive logic. I combined rule-based behaviours with environment-driven inputs, creating an agent that could respond dynamically to queries.
It also pushed me to communicate clearly. Creating a transcript and presentation taught me how to articulate technical ideas for both expert and non-expert audiences. I gained confidence not just in implementation, but in explaining the “why” behind my design decisions.
Completing the research agent gave me a real sense of growth. I now see myself as capable of designing agent-based systems with practical value, while staying grounded in ethical responsibility, technical clarity, and academic rigour.