AI-Powered Listing Filter (MVP Build)

Project Brief – MVP Automation Tool

Overview
We are developing a small, modular MVP tool that will:

Scrape structured data from a public listing website

Store and analyze this data

Rank entries based on basic logic

Send selected entries to a messaging platform (WhatsApp or Telegram)

This is a standalone tool with automation, data handling, and alerting.

Core Components
1. Web Scraper
Scrape key fields (title, price, area, etc.) from a dynamic website

Must handle:

JS-rendered pages

Pagination

Avoiding duplicates

Anti-bot measures (headers, retries, proxies if needed)

2. Data Handling
Store in Airtable, PostgreSQL, or Google Sheets

Auto-update daily with new/changed listings

Exportable CSV or dashboard access

3. Scoring/Filtering Logic
Use formulaic rules (e.g., price per sqft, keyword-based)

Optional: GPT/OpenAI API for enhanced scoring

4. Notification System
Automatically send selected listings via:

Telegram Bot or WhatsApp API (via Twilio)

Format messages nicely with summaries and links

Prefer command-based admin control or minimal UI

5. Automation
Run scraping and alerts on a daily schedule

Use Make.com/Zapier or custom Python scheduler

Ideal Developer
You should be able to:

Build all modules independently

Write clean, well-commented code

Recommend tools when needed

Communicate clearly and respect NDA boundaries

Note
This document contains only the technical overview.
Full specifications and business context will be shared after signing an NDA.