Aaron Gammon

Inferrex - Stop integrating systems. Start understanding them.

byβ€’
Teams stitch together a connector tool, a mapping tool, an MDM tool, and now a GPT wrapper β€” and it still breaks when an API changes. Inferrex replaces all of it with one platform. Its own proprietary AI (InferrexAI, trained on 629k samples) infers schemas, reconciles sources into one governed record, and self-heals when APIs change. At launch: 10,040 API specs across 2,640 vendors, 5M+ cross-provider mappings. SaaS, your cloud, or air-gapped.

Add a comment

Replies

Best
Aaron Gammon
Maker
πŸ“Œ
Hey Product Hunt, I'm Aaron, founder of Inferrex. I spent 18 years selling and implementing enterprise software, and watched almost every organisation lose to the same quiet problem: their systems couldn't understand each other. Everyone chased the "single customer view." Almost nobody reached it β€” because we kept treating it as an integration problem (connect A to B, rebuild it when either side changes) when it's really a comprehension problem. So I built the layer I always wished existed. One platform, end to end: InferrexAI, not a wrapper. Its own proprietary, self-improving model stack β€” no frontier-model dependency, no data leaving your perimeter. Sovereign deployments get a learning loop that improves entirely inside their own security boundary. Schema inference. Point at any API; the schema appears and fields classify themselves in real time. Pipeline graphs. One source, many targets, one pipeline β€” with per-field sync tiers and actions on mappings. Golden layer. Every source reconciles into one authoritative, governed record β€” you always know which system is the source of truth for which field. Self-healing. APIs change; Inferrex detects, re-infers, and repairs β€” silently or with approval. Intelligence. Waste, gaps, and pipeline suggestions surfaced with cost and ROI. Archival & compliance. Signed deletion certificates and UK retention rules, built in. Developer surface. MCP, SDK, CLI, and Copilot β€” drive everything from your IDE. Where it stands at launch: β€’ 10,040 API specs across 2,640 vendors, 35 schema formats (OAS, GraphQL, proto, Smithy, OData…) β€’ 4M+ fields and 647k entities mapped, 5M+ cross-provider mapping pairs β€’ InferrexAI trained on 629k samples, 8 classification layers in production β€’ The platform itself: 1.3M lines generated from 116k handwritten β€” 7.5Γ— codegen leverage, all from manifests (numbers as of launch day β€” the corpus grows with every build) Run it as SaaS, in your own cloud, or fully air-gapped. It's in limited paid pilot ahead of general availability. See it in action: https://inferrex.com/demo Get in touch about a pilot: https://inferrex.com/contact Or just comment here / DM me β€” I'm reading everything today and would love your hardest questions and your worst integration war stories.