Sprint 5 (business path) of Microsoft Discover AI looked at government services: how computer vision and conversational AI can make EV rebate programs easier to run and easier for the public to navigate.
First, a Custom Vision classification / object-detection model to identify whether a parked car is electric — the kind of signal that can trigger automatic rebate-linked discounts without relying on plate lookups alone.
Second, a QnA Maker chatbot trained on EV rebate FAQs so residents get consistent answers and staff spend less time repeating the same guidance.
Together they sketch a lightweight government stack: vision for eligibility signals, NLP for citizen service — built to demo quickly inside the Discover AI format.