Back to products
AI PlatformBeta

BMT AI Platform

Multi-agent orchestration, private RAG, and LLM gateway

BMTAI is BeMind Technology's platform for AI-native products — Python and TypeScript services organized as apps, agents, services, packages, and infra, with unified DevOps.

BMT AI Platform is the shared backbone behind Bemind's AI-native products. Instead of rebuilding agent plumbing for every product, teams compose a handful of focused services — agent coordination, retrieval, an LLM gateway, and payments — that talk over well-defined APIs.

The Oracle coordination layer routes agent-to-agent messages and orchestrates tasks; private RAG grounds every response in your own knowledge base; the LLM gateway routes traffic across providers with fallbacks; and a payments layer handles credits, subscriptions, and x402/Stripe billing — all under one DevOps pipeline.

What you get

  • Multi-agent orchestration and agent runtimes
  • Private RAG over your own knowledge base
  • LLM gateway with provider routing and fallbacks
  • Blockchain and x402-protocol payments, Stripe integration
  • Composable services with unified DevOps across Python and TypeScript

What's inside

The apps, services, and packages that make up BMT AI Platform.

Core services

4

Composable backend services products build on (each is a product in its own right).

Oracle Service
Agent registration, discovery, A2A messaging, and task orchestration
RAG Service
Ingest, chunk, embed, and query your own documents
Knowledge Base Service
Semantic search, Q&A, and real-time content indexing
Payment Service
Stripe payments, credits (FIFO), subscriptions, and invoices

Platform capabilities

5

What the services add up to.

Multi-agent orchestration
Agent runtimes coordinated over real-time transport
Private RAG
Retrieval grounded in your own, isolated data
LLM gateway
Provider routing, fallbacks, and observability
Payments
Blockchain and x402-protocol payments with Stripe
SaaS marketplace
Discover and deploy agents and capabilities

Foundations

4

How the platform is organized.

apps / agents
Product apps and the agents that power them
services
Independent, API-first microservices
packages
Shared Python and TypeScript libraries
infra & DevOps
Unified deployment, observability, and CI/CD