AI·rete·RAG

ai·rete·rag is a decision intelligence platform that pairs a Rete rule engine with retrieval augmented generation. Rules decide the outcome in a deterministic, auditable and repeatable way, and retrieval over the customer's own documents explains the reasoning in plain language. The vendor describes three live patterns. Rules can narrow the scope of retrieval, so a cardiac case fetches only cardiology sources. Retrieval can feed working memory, where documents are parsed into facts such as entities, dates and obligations that rules then act on. Or the engine can decide first and retrieval is used afterward to generate a grounded narrative. Every decision has an audit trail linking it to the rules that fired and those that nearly did, and a policy rules catalog shows conditions, salience and verdicts per domain. The home page offers a no signup demo across loan underwriting, fraud screening and clinical vitals, three of eight built in domains. An MCP integration lets an assistant gather the facts the rules need from a case, with files staying on the assistant side and only facts reaching the API. A public API is included. It is built for developers and teams shipping decision critical products. Plans are priced by decisions per month with no per seat fees. Free is $0 with 1 domain and 1,000 decisions. Supporter is $1 per month, Builder $19, Standard $39, and Pro $119 per month billed annually, with unlimited domains and 500,000 decisions. Enterprise is also listed, and an annual option saves 20 percent on Pro.