Stellenbeschreibung:

Company Description

I run an automotive trading company in Germany . We specialize in buying damaged cars from insurance auctionsand reselling them as they are without repair.


Problem Statement

Currently, I manually evaluate each vehicle based on PDFs from auctions (with photos and value reports and basic car data). I want to automate this process.


Project Goal

The goal is to build a program that can:


Core Functionalities


• Parse Auction PDFs
  Extract car details (brand, model, version, year, mileage, estimated repair cost, market value).
  Process both text fields and attached photos be able to see what parts are damaged and what type of repair is needed.


• Market Price Check
  Automatically search mobile.de and otomoto.pl for the same car (same model, engine version, equipment, year, mileage ±10k).
  Take the lowest dealer offer as the “market value.”
  This might need learning as well? Its hard to evaluate what should be the fair market value, as the damaged cars after repair are around 5% cheaper as normal cars.


• Damage Detection (Computer Vision)
  Analyze auction photos to identify which parts are damaged (front, rear, side, structure, engine, doors, mirrors, airbags etc).
  How this should be fixed by painting, bodywork etc.


• Repair Cost Estimation
  Cross-check auction repair cost.
  Recalculate realistic repair cost by using car part prices provided by us from excel?
  Plus by estimating how much work hors should it be needed
  Allow model learning: the AI can improve by learning from my past manual bids (I can provide data).


• Final Calculation (Business Logic)
  Apply my custom formula:


Optional Features

User interface (web app/dashboard) to upload a PDF and instantly get:
• Car details
• Market value
• Repair cost
• Maximum bid price
• History/log of previous evaluations.

NOTE / HINWEIS:

EN: Please refer to Fuchsjobs for the source of your application

DE: Bitte erwähne Fuchsjobs, als Quelle Deiner Bewerbung