Backend, RPA & AI Developer: WhatsApp Vehicle Auditing Platform - 07/07/2026 21:40 EDT
Budget: $250 – $750 USD
WORKFLOW
1. User sends a vehicle license plate via WhatsApp.
2. Bot processes payment via local gateway.
3. System scrapes 7 Peruvian government portals in parallel and generates a PDF report in under 60 seconds.
4. An interactive AI advisor session opens for a hard limit of 15 questions/messages. The AI acts as a knowledgeable friend or trusted local mechanic—not just discussing price, but natively analyzing the specific vehicle model, brand reliability, performance dynamics in Lima, body style, and offering customized negotiation arguments based on the data.
TECHNICAL SCOPE
We provide 3 JSON schemas, the completed PDF design, and a detailed logic document with the rules to define the exact data that must map into the PDF. No visual design is required.
Parallel RPA Data Engine: Scrape simultaneously using Playwright/Puppeteer + Redis/Celery + Captcha Solver API from:
SUNARP (PAID / Has Captcha): Vehicle specs, ownership data, liens, and previous purchase price from deeds via SPRL login.
SAT LIMA (FREE / Has Captcha): Traffic fines (soles), active capture orders, and vehicle tax debts.
SUTRAN (FREE / Has Captcha): National highway traffic fines (soles) and vehicle retention orders.
MTC CITV (FREE / Has Captcha): Technical inspection status and historical mileage records.
APESEG (FREE / Has Captcha): SOAT status, policy type, and expiration dates.
PNP (FREE / Has Captcha): Active vehicle theft reports and police warrants.
SBS (FREE / Has Captcha): Total loss insurance history and heavy accident alerts.
PDF Compiler: Backend calculates totals, sets a 1-10 Score, maps and injects data based on our provided data rules document into the predefined PDF template, and sends it via WhatsApp.
AI Engine & Advanced Knowledge Base (LLM): Post-delivery chat window. The prompt dynamically injects:
JSON 1: Scraped report and vehicle valuation data.
JSON 2 (Proprietary Intelligence Data Engine): A high-value database containing deep technical, safety, and market metrics for 150 car models. This engine is fueled by a comprehensive bi-annual scheduled scan/scraping of external automotive authority platforms (e.g., NeoAuto, Latin NCAP, OLX Autos, and official brand recall databases). This database acts as a proprietary knowledge mine, supplying critical safety scores, known manufacturing defects, and structural metrics directly into the final PDF report.
JSON 3: Chat history (user/assistant roles) and a strict 15-message countdown counter.
Conversational & Negotiation Logic: The AI acts as a peer expert fluent in local Peruvian car context. It analyzes vehicle style, durability, and performance for Lima's unique traffic/terrain. If the user types a street price, the system calculates the gap against our server valuation, and the AI matches it with structural defects or debts to formulate natural counter-offers.
UX Guardrail: Tracks user intention. If frustration or a refund is requested, it replies with an empathetic message and provides the official support email.
Admin Dashboard & Data Monetization Analytics: A centralized back-office panel optimized to aggregate high-value market insights:
Transactions & Logs: Track active orders, payment statuses, PDF delivery, and user feedback (Likes/Dislikes).
RPA Health Status: Real-time uptime/timeout monitor for all 7 external portals.
CRUD Dictionary: Simple UI to update the 150-car local intelligence database.
Commercial Monetization Metrics: Aggregated analytics showing top-searched brands/models (market demand share), average volume of traffic debts in soles by vehicle category/district, and automated alerts for "high-speculation flipping vehicles" (units with recent ownership transfers under 90 days).
Automated Recovery Loop: Implements background jobs to trigger a native WhatsApp follow-up message if a user initiates a query but drops off before completing the payment gateway transaction.
TECH STACK
Backend: Python (FastAPI), Node.js, or Go.
RPA: Puppeteer or Playwright + Captcha Solver.
Queue: Redis / Celery.
DB: PostgreSQL or MySQL.
API: Meta Cloud API / Twilio + Payment Gateway.
To Apply: Send your portfolio showing high-speed concurrent web-scraping with captchas, WhatsApp state management, or LLM context integration. No generic templates.
1. User sends a vehicle license plate via WhatsApp.
2. Bot processes payment via local gateway.
3. System scrapes 7 Peruvian government portals in parallel and generates a PDF report in under 60 seconds.
4. An interactive AI advisor session opens for a hard limit of 15 questions/messages. The AI acts as a knowledgeable friend or trusted local mechanic—not just discussing price, but natively analyzing the specific vehicle model, brand reliability, performance dynamics in Lima, body style, and offering customized negotiation arguments based on the data.
TECHNICAL SCOPE
We provide 3 JSON schemas, the completed PDF design, and a detailed logic document with the rules to define the exact data that must map into the PDF. No visual design is required.
Parallel RPA Data Engine: Scrape simultaneously using Playwright/Puppeteer + Redis/Celery + Captcha Solver API from:
SUNARP (PAID / Has Captcha): Vehicle specs, ownership data, liens, and previous purchase price from deeds via SPRL login.
SAT LIMA (FREE / Has Captcha): Traffic fines (soles), active capture orders, and vehicle tax debts.
SUTRAN (FREE / Has Captcha): National highway traffic fines (soles) and vehicle retention orders.
MTC CITV (FREE / Has Captcha): Technical inspection status and historical mileage records.
APESEG (FREE / Has Captcha): SOAT status, policy type, and expiration dates.
PNP (FREE / Has Captcha): Active vehicle theft reports and police warrants.
SBS (FREE / Has Captcha): Total loss insurance history and heavy accident alerts.
PDF Compiler: Backend calculates totals, sets a 1-10 Score, maps and injects data based on our provided data rules document into the predefined PDF template, and sends it via WhatsApp.
AI Engine & Advanced Knowledge Base (LLM): Post-delivery chat window. The prompt dynamically injects:
JSON 1: Scraped report and vehicle valuation data.
JSON 2 (Proprietary Intelligence Data Engine): A high-value database containing deep technical, safety, and market metrics for 150 car models. This engine is fueled by a comprehensive bi-annual scheduled scan/scraping of external automotive authority platforms (e.g., NeoAuto, Latin NCAP, OLX Autos, and official brand recall databases). This database acts as a proprietary knowledge mine, supplying critical safety scores, known manufacturing defects, and structural metrics directly into the final PDF report.
JSON 3: Chat history (user/assistant roles) and a strict 15-message countdown counter.
Conversational & Negotiation Logic: The AI acts as a peer expert fluent in local Peruvian car context. It analyzes vehicle style, durability, and performance for Lima's unique traffic/terrain. If the user types a street price, the system calculates the gap against our server valuation, and the AI matches it with structural defects or debts to formulate natural counter-offers.
UX Guardrail: Tracks user intention. If frustration or a refund is requested, it replies with an empathetic message and provides the official support email.
Admin Dashboard & Data Monetization Analytics: A centralized back-office panel optimized to aggregate high-value market insights:
Transactions & Logs: Track active orders, payment statuses, PDF delivery, and user feedback (Likes/Dislikes).
RPA Health Status: Real-time uptime/timeout monitor for all 7 external portals.
CRUD Dictionary: Simple UI to update the 150-car local intelligence database.
Commercial Monetization Metrics: Aggregated analytics showing top-searched brands/models (market demand share), average volume of traffic debts in soles by vehicle category/district, and automated alerts for "high-speculation flipping vehicles" (units with recent ownership transfers under 90 days).
Automated Recovery Loop: Implements background jobs to trigger a native WhatsApp follow-up message if a user initiates a query but drops off before completing the payment gateway transaction.
TECH STACK
Backend: Python (FastAPI), Node.js, or Go.
RPA: Puppeteer or Playwright + Captcha Solver.
Queue: Redis / Celery.
DB: PostgreSQL or MySQL.
API: Meta Cloud API / Twilio + Payment Gateway.
To Apply: Send your portfolio showing high-speed concurrent web-scraping with captchas, WhatsApp state management, or LLM context integration. No generic templates.
Related categories:
Python
Web Scraping
MySQL
Node.js
PostgreSQL
Redis
API
Celery
FastAPI
RPA Development