BillyCheck
SaaS application for energy bill optimization. Upload, AI analysis, provider comparison, savings in 30 seconds.
From idea to product in production2025
A real problem nobody was solving simply.
In Belgium, energy bills are opaque. Rates vary by region, meter type and distribution network operator. Comparing offers requires understanding complex rate structures that even informed consumers don't master. Existing simulators are slow, technical, and require filling in dozens of fields manually. BillyCheck had to solve this problem in one action: take a photo of your bill and immediately see how much you can save.
Turning a complex process into a 30-second experience.
The challenge wasn't technical in the classic sense. It was a product challenge. How do you reduce a process that normally takes 15 minutes of manual input and knowledge of the Belgian energy market to a 30-second experience? How do you extract reliable data from a bill photo, match it against a real pricing model with regional variables, and display a credible result? And how do you do all of this without login, without friction, without a tutorial?
Thinking in product, not in features.
We started with the end result: the screen the user sees after scanning their bill. That screen decides whether the product has value or not. We designed backwards: first the results experience, then the upload flow, then the technical architecture to support both. The key decision was eliminating mandatory login. A user discovering the product should be able to scan a bill and see their savings without creating an account. Authentication comes after value, not before. This choice conditioned the entire technical architecture.
A three-step flow. Zero friction.
The interface was designed around a linear journey: upload, analysis, results. No complex dashboard. No menu to figure out. The user arrives, uploads their bill (photo or PDF), and AI does the rest.
Data extraction relies on GPT-4o Vision. The model analyzes the bill, identifies relevant fields (consumption, rate, distributor, meter type, region) and structures them for the comparison engine. We built a pricing model that integrates Belgian market specifics: regional rates, GRD codes, single and dual-rate structures, variable distribution costs by zone.
The comparison engine matches extracted data against offers from 15+ Belgian energy providers. The result displays potential savings by provider, ranked by relevance.
The business model is freemium with Stripe. The first scan is free. Subsequent scans are available via paid credits.
Full-stack. From design to deployment.
The application was built in Next.js with a PostgreSQL database via Prisma. Authentication works via cookies, with no mandatory login for the first scan. GPT-4o Vision integration handles billing data extraction. The pricing engine was custom-built for the Belgian energy market. Payment runs through Stripe with a credit system. The entire product was designed, built, developed and deployed by P-XEL. One studio. From idea to product in production.
A product that works. Not a mockup.
BillyCheck is in production. Users scan bills, get comparisons and identify savings. The product does in 30 seconds what used to take 15 minutes of manual research. The experience is smooth, the AI is reliable, the monetization model is in place. This is exactly what "we design, we build, we ship" means.
- Complete product design (UX research, user journeys, UI)
- Next.js web application in production
- GPT-4o Vision integration for bill extraction
- Belgian energy pricing model (15+ providers)
- Stripe payment system (freemium + credits)
- PostgreSQL / Prisma database
- Deployment and production launch
A digital product to design and ship?
We design web applications and MVPs from idea to deployment. One studio. Zero middlemen. web applications and MVPs