Tampa Energy Intelligence

Inspiration:

Cities are making million-dollar energy decisions using spreadsheets and gut instinct. We wanted to build a tool that gives any city planner instant visibility into where their energy is going, what to fix first, and how today's decisions connect to a 30-year sustainability vision — before the next budget meeting, not after.

What it does:

Tampa Energy Intelligence lets city planners upload their district energy data and instantly see a live interactive map of Downtown Tampa color-coded by waste level. An AI recommender ranks every sustainability project by a composite score weighing efficiency, cost, speed, and disruption. A full analytics dashboard shows consumption trends, peak demand patterns, and power budget breakdowns. A multi-horizon roadmap connects today's actions to net-zero Tampa by 2054. And a Gemini-powered AI assistant lives on every page — screen-aware, always on, answering questions with full knowledge of what the planner is looking at.

How we built it:

We built Tampa Energy Intelligence with Next.js, TypeScript, Tailwind CSS, Leaflet.js, Recharts, Google Gemini 2.0, and Framer Motion. Gemini powers the AI assistant, page context injection, and real-time data-aware responses. Leaflet.js renders the interactive Tampa district map with live polygon overlays. The recommender engine scores every project using a weighted formula: efficiency gains times savings divided by cost times implementation time times disruption. The platform reads the user's current page state — selected district, active map layer, visible KPIs — and injects that context into every Gemini API call so the assistant always knows what the planner is seeing.

Challenges we ran into:

The hardest part was making the AI assistant genuinely useful rather than generic. A chatbot that answers energy questions is easy. A chatbot that knows you are currently looking at Downtown Core, that the active layer is Peak Demand, and that three alerts are flagged — and answers accordingly — required building a real-time context collection system that reads live page state on every message. We also had to design the recommender scoring formula carefully. Weighting efficiency, cost, speed, and disruption in a way that surfaces genuinely actionable projects — rather than just the cheapest or the most impactful in isolation — required iteration to get the balance right.

Accomplishments that we're proud of:

We are proud that Tampa Energy Intelligence is more than a dashboard. It creates a full decision-support environment — interactive map, AI recommendations, analytics, multi-horizon roadmap, and a conversational assistant — that any non-technical city official can use on day one without training. We are especially proud of the screen-aware AI assistant. It does not just answer questions about energy in general. It answers questions about the specific district the planner clicked, the specific KPIs on their screen, and the specific alerts flagged in their data. We are also proud that every recommendation is fully explainable. The city can defend any spending decision because the scoring formula is visible, the inputs are real, and the logic is transparent.

What we learned:

We learned that the hardest part of a city-facing AI product is not the AI — it is the workflow fit. The platform only works if it mirrors exactly how a city planner actually thinks: get the data in, understand what it shows, prioritize what to fix, connect it to a long-term plan. Every feature we cut or kept came back to that question. We also learned how critical context injection is for making AI assistants actually useful. Without knowing what page the user is on, what district they selected, and what data is visible, the assistant gives generic answers. With that context, it gives answers that feel like talking to an expert who was in the room the whole time.

What's next for Tampa Energy Intelligence:

The immediate next step is connecting the platform to real city energy data through utility API integrations, replacing mock data with live feeds from Tampa Electric and municipal building management systems. After that, expanding beyond Downtown Tampa to all of Hillsborough County, then positioning as a deployable SaaS product for mid-size cities nationally. The long-term vision is a platform where every city in America can upload their energy data and get a credible, AI-generated path to net-zero — in minutes, not months.

Built With

• python