Hi Product Hunt 👋 I’m Garfield, building Assay. Most enterprise AI problems aren’t only model problems. Companies have years of duplicated, stale, ownerless, and conflicting material spread across wikis, drives, chat, and business systems. When that material goes directly into AI, the AI can become confidently unreliable. Assay is our attempt to build the knowledge supply chain between source systems and enterprise AI applications and Agents. It connects read-only, preserves permissions, identifies exact and near duplicates, applies explicit survivorship rules, rates trust using freshness, source authority, and human verification, and keeps lineage from source to AI use. This is an early-stage product and interactive demo—not a claim of a finished enterprise deployment. You can explore the product or take the free two-minute knowledge readiness assessment to see which governance dimensions need attention first. I’d especially value feedback on: • Which knowledge quality failure causes the most pain in your AI or RAG work? • Would you start with deduplication, trust rating, permissions, or human review? • What would you need to see before connecting a source read-only? Thanks for taking a look.