Features

Stop guessing what your AI just did

New nodes, 300+ tools, and drop-in API compatibility mean you build, connect, and deploy multi-agent workflows in days - not another roadmap cycle.

01
TRIGGER
input.received
webhook
02
AGENT
plan.compose
reasoning
03
TOOL
web.search
live data
04
LOOP
iterate.until(done)
×2
05
EGRESS
response.sent
traced
explainable run · 5 stepsevery step logged
01
ORCHESTRATE

Every Moving Part Your Agents Need, Already Built In

01.01

Voice Agents

Replace IVRs and scripted bots with intelligent Voice Agents that understand context, remember past interactions, and respond naturally.

01.02

130+ New Tools

A large library of pre-built integrations, part of the growing 300+ tool ecosystem built to extend what your workflows can do.

01.03

Smart Web Search

Context-aware web search that pulls live, up-to-date information into workflows on demand, keeping every response grounded in reality.

01.04

Loop Node

Iterative task execution until a condition is met - true programmatic control over workflows, no code required.

02
TRAIN

Stop Deploying AI That Has to Guess

02.01

Model Training

Build private AI models using your organization's own knowledge, data, and expertise - fully in-house.

02.02

RAG

Connect AI to enterprise data sources for accurate, source-backed responses grounded in real context.

02.03

BM Architect

Design, test, and refine AI agents with visual workflows before deployment, all from one governed portal.

03
CONNECT

Every System You Already Run Is Already Welcome Here

03.01

Webhooks

Connect Blue Mesh to any external system with incoming webhook endpoints that receive and process real-time events as they happen.

03.02

Custom API Connectivity

Allows external APIs to be plugged in as reusable tools, extending what workflows can access and do.

03.03

OpenAI-API Compatibility

Integration of studio deployments as OpenAI APIs, letting existing workflows run without any modification.

03.04

OpenRouter Connectivity

Enables direct integration with OpenRouter to access multiple AI models through a single, unified API.

04
OPERATE

Explainable, Accountable, and Never a Surprise on the Invoice

04.01

Explainable AI

Displays real-time, step-by-step agent actions including reasoning, tool usage, and execution status for full transparency.

04.02

Token Tracking

Track and monitor token consumption across every request in real time, giving teams full visibility to control cost and optimize usage.

Where to go next
  • BM Studio The canvas these nodes, tools and loops are assembled on, drag and drop, with no code.
  • BM Architect Describe a workflow in plain language and have it designed, tested and deployed, with a demo to show for it.
  • BM Oasis Fine-tuning on your own data, and a straight answer on who keeps the weights that come out of it.
  • BM in a Box All of it as a pre-configured appliance on NVIDIA DGX hardware, running with no outbound connection.

Unlock voice AI at scale with an API Call

Bring a real workflow. In one session we will build an agent for it, live, and show you exactly where your team stays in control.