Calculator Timeless
Built an AI quote calculator for a tree service business to turn incomplete inquiries into structured, actionable leads with instant pricing feedback.
Overview
The business needed a faster way to qualify leads and give customers immediate pricing clarity without long back-and-forth. I designed and shipped a multi-step calculator that captures job-critical details, generates an AI-assisted quote, and sends structured submissions to the admin inbox. I also handled deployment and debugging across Wix embed constraints, CORS behavior, and lead-delivery reliability. The final build improved intake quality and made quotes usable as a real first-pass price conversation.
Business Problem
Before this calculator, quote intake had three issues:
- Leads came in incomplete (missing size, access, debris preferences, etc.)
- Response time was inconsistent because staff had to ask follow-up questions
- Pricing expectations were unclear, causing back-and-forth before site visits
Solution Implemented
I created a guided multi-step form with AI-assisted pricing logic and automated lead delivery.
Key features:
- Step-by-step questionnaire for service type, tree count, height, trunk size, species, access, condition, obstacles, and debris handling
- Photo upload support for better estimate quality
- AI-generated line-item pricing and a single final customer-facing price
- Immediate quote display after form completion
- Lead submission to Web3Forms with all customer + project details
- Mobile-friendly UX tuned for embedded Wix deployment
Technical Approach
The solution uses a lightweight frontend flow and an API-backed AI layer.
- Frontend: Single-page HTML/CSS/JS wizard experience
- AI endpoint: Cloudflare Worker proxy for AI quote generation
- Lead handling: Direct Web3Forms submission from client-side form data
- Platform integration: Embedded in Wix via HTML embed widget
Major Iterations & Fixes
During implementation, I improved both reliability and quote usability:
- Removed auto-advance on single-choice answers so users must click Continue deliberately
- Switched quote output from broad “range” toward a more actionable fixed-price result
- Added fallback normalization so older AI range formats can still map to one usable number
- Removed “Palm Tree Service” as a standalone option to avoid category overlap
- Refined visual styling to a clean white/tan/green brand palette
- Debugged CORS + deployment cache issues specific to Wix embedded contexts
- Improved lead submission diagnostics to reduce silent failures
What This Project Demonstrates
This project highlights practical skills in:
- AI prompt engineering for business outcomes
- Frontend UX workflow design for conversion flows
- API integration + error handling under real deployment constraints
- Iterative product improvement based on live testing feedback
What we shipped
◆ Designed a step-by-step intake flow that captures service scope, tree specs, site access, risk factors, and debris preferences.
◆ Implemented client-side validation and manual step progression to improve completion control and reduce accidental skips.
◆ Integrated a Cloudflare Worker endpoint for AI quote generation and hardened prompt/output handling for consistent pricing structure.
◆ Refined quote presentation from broad ranges toward a single actionable price format with fallback normalization logic.
◆ Connected form submissions to Web3Forms, then debugged and stabilized delivery behavior across Wix embed and API constraints.
◆ Removed overlapping service options and aligned category logic so pricing inputs stayed clean and non-duplicative.
Tools, standards, assumptions.
- HTML
- CSS
- JavaScript
- Cloudflare Workers
- Anthropic API
- Web3Forms
- Wix
- Wix
- AI Quote Engine
- Claude API
- Web3Forms
- Cloudflare workers
What changed.
Numbers from the live store after the work was shipped — measured against pre-existing baselines, never theoretical.