The problem statement: "Buy or Wait?"
Build an AI financial agent that answers one question — can this user actually afford this expense right now?
Sounds simple. It wasn't.
The agent couldn't just check a balance. It had to reason through recurring expenses, pending payments, essential spending, confirmed income — even details buried in messages and images.
For every request, it had to decide: pay in full, pay partially, use installments, wait, or don't proceed at all.
Here's what made it hard — the recommendation had to be personal.
Two users. Same balance. Different financial habits, different priorities, different answers.
So the agent didn't just calculate a number. It built a payment plan. Flagged which expenses to cut. Explained the reasoning behind every decision.
24 hours to design, build, and submit.
Then a 30-minute live interview with an AI Judge — camera on, walking through my architecture, my decisions, and how I actually used AI throughout the build.
A completely different kind of evaluation. Not just "does the code run" — but "do you understand what you built."
Big learning on personal finance AI, agentic decision-making, and building under real time pressure.
SO, Do Not forget to follow to get free upcoming Resources to Boost your AI Capabilities
Thanks 🙌❤️
Top comments (0)