I need a developer to build a reliable Keepa API tool that can lookup Amazon product data via barcode, ASIN, or title, and return structured results. The goal is to automate product validation for Amazon FBA/wholesale sourcing. Please only apply if you have real Keepa API experience and understand token efficiency, and title matching. Overview The tool needs to run in two modes: Batch Mode • Input: CSV containing product info and some or all of the following fields: ean/upc, asin, or title • Output: results csv (original data + Keepa results) Single-Item Mode • Called from another script via command line • Accepts ean/upc, asin, or title • Returns one JSON object to stdout so another script can read it Matching Requirements This is the most important part of the job. Previously I've had challenges at reliable title matching. I want to know: How will you accurately match products by title when barcode is missing or incorrect? You may use strict matching, fuzzy matching, token matching, or a hybrid — I just need something reliable. Required matching behaviour: • Try barcode first • If barcode fails or is missing, match using title • If multiple possible matches exist, put the row in errors.csv with match confidence • The script must allow selecting which matching pass to run (barcode only, title, fuzzy, all, etc.) • In batch mode, the script should complete barcode matching for the full CSV first, then move on to the next match type, instead of trying all match methods per row. • If an ASIN is provided, no matching logic is required — just fetch Keepa data for that ASIN Keepa Data Needed For each matched ASIN, return the following: ASIN Title Brand Main category rank 30-day average rank 90-day average rank Current Buy Box price 30-day average Buy Box price 90-day average Buy Box price Lowest new price Amazon price (if Amazon is selling it) Whether Amazon Retail is a seller FBA offer count FBM offer count Timestamp Marketplace Control Primary marketplace is Amazon UK. The script must accept the domain/marketplace as a parameter (e.g., UK, US, DE, etc). Token Efficiency & Performance This matters because Keepa charges tokens per request. A good implementation will include things like: • De-duplicate lookups so ASIN/EAN/title isn’t queried multiple times • Use caching to avoid unnecessary re-requests • Batch Keepa calls • Respect Keepa rate limits and token refill time • Retry/backoff on 503 or token exhaustion • Efficient matching pipeline You can implement this however you think best — I just need it to be token-efficient and reliable. The output should update a partial output file every 50 records (or something similar). The script should output progress to the console during execution. Deliverables • Python-based tool • Command-line interface • Single-item mode JSON output • Clean and simple setup instructions • A short explanation of your title matching approach • Example output from similar Keepa work if available To Apply, Please Answer: Do you have prior experience with the Keepa API? How would you approach accurate title matching when barcodes aren’t reliable? How do you typically ensure token-efficient batch processing? Estimated timeframe and cost. Any examples of similar work. Notes I am technically competent, so please be direct and specific. I will test the tool with large CSVs, missing barcodes, mismatched titles, and edge cases. Thanks, Tim