Back to the website
A WORKING PRODUCT WALKTHROUGH

Explore a real rules engine.

Choose a fictional client, adjust sample balances and check off actions. This browser-only demo uses the same calculations and recommendation engine as the authenticated application.

Synthetic data only. No API calls or saved records.
01 / FICTIONAL ASSESSMENT

Getting rental-ready

A fictional renter applying soon, with high revolving usage.

Illustrative revolving utilization75.0%
Paydown to the illustrative combined 29% target$920.00Not a score prediction; protect rent, necessities and required payments first.
The app computes all financial figures itself. AI, when separately enabled and consented to in the private workspace, can only explain approved actions.
02 / DETERMINISTIC ACTION PLAN0/4 done

Your next steps

Each recommendation below is generated from the sample inputs without a language model.

03 / CLIENT-CONTROLLED COLLABORATION

What would a Realtor see?

Only when a client explicitly shares: a chosen label, the program, and aggregate checklist progress. No account balances, scores, explanations or itemized actions.

Educational planning only. This demo cannot estimate score changes, lending eligibility, rental screening decisions or approval outcomes. Never use real client information in these sample controls.

This portfolio walkthrough uses fictional inputs. The private product adds Supabase authentication, saved assessments and plans, activity history, and consent-based OpenAI personalization. Those features require backend configuration and independent live verification.

See the private workspace