10 AI Agent ideas. One is yours to build.
Your Week-4 capstone is one AI Agent, built individually. The cohort works from a set of 10 beginner-friendly agent ideas. Each follows the same backbone, but the agent focus changes, so every portfolio is distinct.
Suggests career paths and a learning roadmap from a student's interests and skills.
RAG · prompting · Gemini API · profiles
Generates role-specific questions and coaches better answers.
Prompt patterns · roles · feedback loops
Reviews a resume and suggests targeted, practical improvements.
Structured output · prompting · file input
Answers student questions from a college's own information using RAG.
RAG · embeddings · retrieval · grounding
Turns goals and deadlines into a realistic weekly study plan.
Reasoning · structured output · scheduling
Turns raw notes or transcripts into clean, structured study notes.
Summarisation · prompting · files
Categorises spending and offers simple, friendly budgeting tips.
Data handling · prompting · categorisation
Gathers, summarises, and organises sources on any topic.
RAG · summarisation · tools · memory
Builds a step-by-step roadmap to learn any new skill.
Reasoning loops · structured output
Plans tasks, sets reminders, and drafts quick replies.
Agents · tools · memory · automation
The agent backbone: a clear problem statement, a design, an implementation in Python that calls the Gemini API, prompt engineering and where useful RAG over your own documents, a GitHub repository with a README, a 3 to 5 minute demo video, and a live Demo Day presentation.
A working AI builder's toolkit
Learn on free tools and leave with the core GenAI skills employers look for.
Ready to build your portfolio?
Join the cohort, ship your own AI Agent, and demo it on Demo Day.