SaleClarity

• 

• 

• 

• 

• 

• 

• 

• 

Inspiration

SaleClarity grew from my conveyancing studies and work on two property transaction tools: Properoute and Completia.

Preparing to sell a property can leave people unsure what information to gather, which concerns to raise and how to keep everyone informed. I wanted to give sellers a dedicated starting point: practical guidance that helps them understand their circumstances, act sooner and keep track of what comes next.

The PayPal AI Hackathon inspired me to connect that guidance to useful digital support that the seller can understand and choose.

What it does

SaleClarity asks a short sequence of questions about the seller’s stage, needs and concerns. Questions advance as they are answered, keeping the experience manageable.

AI turns those selections into a personalised explanation and three practical next actions. Defined matching rules identify whether Properoute, Completia or both could help, with a clear reason for the recommendation.

Properoute helps sellers organise property information and prepare a handover record. Completia supports clear transaction progress updates between the parties.

When a seller identifies a lease or property concern, SaleClarity asks whether their solicitor has already been told. The guidance then helps them raise the concern or track the response, leaving legal conclusions to their solicitor.

The action plan appears before payment. Sellers can change their answers, receive revised guidance or use the guidance without purchasing.

If they choose a recommended app, SaleClarity shows the price and access details before opening PayPal Sandbox checkout. A completed test payment activates the selected browser-session demo pass. Its purchase button becomes an Open button, and both passes remain visible when both products are purchased.

With the seller’s agreement, their action plan can also appear inside Properoute.

How I built it

I developed SaleClarity with ChatGPT and Codex assisting the design, implementation and testing. The interface uses HTML, CSS and JavaScript. The hosted application runs through OpenAI Sites on Cloudflare Workers.

Server-side requests to the OpenAI API generate structured guidance containing an explanation and three actions. Product selection, prices and navigation remain controlled by the application. The API key stays on the server.

PayPal’s JavaScript SDK provides Sandbox checkout. The application checks the returned capture status and payment amount before recording the selected demo pass in browser-session storage.

Cloudflare D1 stores usage counters so public-demo AI limits persist across server restarts. Requests reserve their allowance before calling the model.

The mobile interface includes a prominent working notification with animated dots and a gentle fade-in when guidance arrives. Reduced-motion preferences are respected.

The public repository includes the SaleClarity assistant and purchase journey, a local Node.js and SQLite runner, testing instructions and automated checks. Properoute and Completia remain separate connected products.

Challenges

The main challenge was making the connection between AI and payment useful. Sellers need practical help and a reason to consider a product before being asked to buy it.

I placed the guidance before the purchase options, simplified the questions and kept purchases optional. I also separated product-matching rules from AI-generated wording so recommendations remain consistent and testable.

Another challenge was preserving a clear experience after purchase. Each app now has one card whose action changes from Buy to Open, avoiding duplicate buttons while keeping both activated products accessible.

On mobile, the initial AI loading indicator was too easy to miss. I moved it into a visible notification so sellers can see that their request is being processed without scrolling to the output area.

Property transactions also require careful boundaries. SaleClarity helps sellers organise facts, understand next steps and formulate questions. It does not verify title, provide legal advice or guarantee a sale.

Accomplishments that I’m proud of

I brought together a working journey from guided questions to personalised AI actions, optional product selection, PayPal Sandbox checkout and activated app access.

The assistant provides useful guidance before any purchase. It can recommend both tools when both needs are present, while explaining their different purposes.

I also made the demo publicly accessible without a ChatGPT sign-in, added persistent usage controls and prepared a runnable public repository. The repository’s automated checks cover recommendation rules, follow-up questions, usage limits and key interface behaviour.

What I learned

I learned that AI becomes more useful when it produces a small, relevant result. Three practical actions can give a seller a clearer starting point than a long explanation.

I also learned that product recommendations need consistent rules, while AI can make the guidance more personal and understandable.

Integrating PayPal taught me to distinguish a completed payment demonstration from a production licensing system. Showing checkout is only part of the journey; the seller also needs clear confirmation and an obvious way to open what they selected.

Testing on smaller screens reinforced how easily an important signal can be missed when it sits outside the user’s immediate view.

What’s next

The next priorities are direct plan import into Completia, stronger continuity between the connected apps and further improvements informed by seller feedback.

The current build uses PayPal Sandbox and browser-session demo passes. Production development would add server-side payment verification, durable licence records and reliable access recovery.

The aim remains the same: a clear, approachable assistant that helps sellers work through their circumstances, act sooner and choose useful support.

Built With

• cloudflare
• codex
• css3
• html5
• javascript
• openai
• paypal
• workers