Ploutos

Auth0 Login
Sandbox Preview
Optimization Map
Optimized Property Rating

Inspiration

The problem starts before the business does. Thousands of Canadian entrepreneurs never launch because the financial complexity stops them before they start and the ones who do launch are often making six-figure decisions completely blind. 98% of Canadian businesses are SMBs, and whether you're a first-time founder or a shop owner on King Street, the tools that exist were built for accountants, not for you. They deserved something better.

What it does

Ploutos gives upcoming and small business owners instant financial clarity without spreadsheets or jargon. Upload your existing financials, watch them map into a live visual node graph, and drag and drop your input costs and set revenues. An AI Accountant Agent powered by Gemini scores your Business Health out of 100 and flags real risks. When you're ready to expand, our search engine pulls live foot traffic from BestTime, competitor density from OpenStreetMap, and interest rate data from the Bank of Canada, then scores locations, visible an interactive 3D map so you know exactly where to open next and why.

How we built it

Frontend: Next.js 14, Tailwind CSS, Mapbox GL JS for 3D location mapping Backend: Python FastAPI serving the AI pipeline AI: Gemini via Backboard for RAG-powered financial analysis and location scoring Data: BestTime API for foot traffic, Overpass/OpenStreetMap for competitor density, Statistics Canada for demographics Auth: Auth0 for secure user authentication, google 0Auth 2.0

Challenges we ran into

Finding reliable API's was the most challenging, as a lot of our project relies on taking data from both external sources and directly form the user to form some output. Also creating both the map and sandbox type of infrastructures were new to the whole team but we tackled the problem head on by seeing how it works and how other hackers in the past have implemented it. We also were targeting backboard as we thought that their technology would be a good fit to what we are doing, but it was also a steep learning curve.

Accomplishments that we're proud of

The end-to-end pipeline actually works, a user can go from uploading a spreadsheet to seeing AI-scored expansion locations on a 3D map in under 30 seconds. The design system we built feels genuinely premium for a hackathon project. And the location scoring combining real foot traffic, competitor data, and macroeconomic signals is something we haven't seen in any tool built for SMBs or upcoming business' .

What we learned

Free APIs have real limits that hurt at scale — architecture decisions like pre-baking coordinates instead of live geocoding made the difference between a broken demo and a working one. We also learned that RAG context quality matters more than model choice — the richer the location documents we fed Backboard, the sharper Gemini's recommendations became. Building for non-technical users means hiding enormous complexity behind a single drag.

What's next for Ploutos

Ploutos isn't able to grab real time data for commercial real estate listings, ideally we would use the realtor.ca's API or other API's that can give us access to this live data
Using more API's for optimization accuracy, such as Google Places API, and using more API's for more diverse inventory/supply items, include tariff tracker to update supplies -Create projection plans for business' to follow based off finances and sandbox adjustments, sales velocities with square and inventory automation -Using tools are larger scale