Claude Code Automotive Product Catalog Data Specialist (ACES/PIES)

*45 Hour+ Work Week, NEW YORK Time Zone USA* This is a full-time, long-term specialist role. Your sole focus is owning our aftermarket car parts catalog end to end. You must be genuinely fluent in Claude Code and Claude in Cowork — this role runs on AI-driven data pipelines you'll build, not manual row-by-row editing. If you don't already work this way, this isn't the role for you. First, what this job actually is (read this even if you've never sold car parts): In auto parts, fitment is everything. A customer needs to know a part fits their exact year, make, model, engine, and trim — order the wrong one and it gets returned. Manufacturers publish their product data in standardized industry feeds (you'll hear them called ACES — the fitment/application data, what fits what — and PIES — the product info: descriptions, attributes, images, pricing). You don't need to be an ACES/PIES expert on day one. You'll grab the manufacturer's XML files, drop them into a conversion tool we already have, and start from there. The real job is what happens next: sanitization. Manufacturer data is almost never clean. You'll constantly see things that don't make sense — garbled titles, missing attributes, inconsistent naming, duplicate or contradictory entries, codes where words should be. Your job is to fix it with judgment, and that means web research is core to the role: when the source data is wrong or unclear, you research the actual part, check how the leading competitors present it, and write it up better — clearer, more complete, and structured around how a real buyer searches and decides. The end goal: we don't just want to display a catalog. We want the RIGHT, ready-to-read, end-consumer catalog — so a shopper can instantly understand the product and confidently buy the correct part for their car. What you'll own: Extract & sanitize. Pull manufacturer ACES/PIES feeds, run them through our conversion tool, then clean and normalize the output into ingest-ready data — validated, deduplicated, consistent. Research & fix. When data doesn't make sense, research the real part and how top competitors present it, then make ours clearer and better based on buyer behavior. Content quality. Ensure titles, subtitles, descriptions, FAQs, and specs are accurate, human-readable, and written to convert — not just copied from the raw feed. Monitor & sync to RDS. Track manufacturer updates, generate clean diff/update files, and build an idempotent pipeline that syncs into our MySQL RDS and loads reliably without breaking existing records. Kit generation. Build kits by finding overlapping applications and matching fitment specs across SKUs. Image pipeline (S3). Source and push product images to S3, correctly mapped to the right product/variant — no orphans, no mismatches. Gap analysis & multi-channel. Identify brands and SKUs competitors carry that we don't, and maintain a pipeline that syncs the catalog to RDS and out to our eBay stores via the eBay listing API. Who you'll work with: You'll work directly with the founder, who is a senior Claude Code prompt engineer and has built the entire system himself — the tooling, the pipelines, the architecture. You won't be handed vague tickets; you'll architect the pipeline flow with him and execute it. That means you need to actually think in Claude Code, move fast, and hold a high quality bar. You should have: Strong, demonstrable fluency with Claude Code and Claude Cowork — building and running data pipelines, not just chatting with an AI Excellent research instincts and judgment — you can look at bad data, figure out what's right, and fix it Sharp eye for content and consumer-readiness — you understand how a buyer reads a product page and what makes them trust and buy Comfort with large XML/CSV feeds, MySQL/RDS, S3, and ETL/sync workflows SQL and scripting (Python preferred) Obsessive attention to detail — you catch the broken fitment, the bad price, the missing image before it ships Bonus: prior auto parts / ACES/PIES exposure, or eBay/marketplace listing experience Details: Long-term, You'll work independently but in close sync with the founder