What does this do?
This tool uses OpenAI (GPT-4o-mini) to analyze and summarize lengthy inspection reports from auction sites. When you import auction data with detailed inspection reports, the raw condition notes can be extremely long. This AI summarization creates detailed, structured summaries (up to 2000 characters) that focus on key valuation factors: overall condition, mechanical issues, hydraulics, undercarriage, structural damage, and operational status.
Storage: Condition notes can now be up to 10,000 characters (previously 2000).
Cost: ~$0.15 per million input tokens. For 100 records with 5000 chars each, that's about $0.08.
Test Single Summary
Paste an inspection report HTML or text to see how the AI summarizes it:
Batch Process Auction Records
Process auction records with long condition notes (>1000 chars). This will:
- Find up to 100 records with condition notes longer than 1000 characters
- Send each to OpenAI for detailed analysis and summarization
- Update the database with structured summaries (up to 2000 chars)
- Preserves critical details for accurate valuation
- Process takes about 1-2 seconds per record (includes 1s delay to avoid rate limits)
Warning: This will overwrite existing condition_notes. Make sure you have a database backup before running batch processing.
Processing records... This may take a few minutes.