Coming soon

BoBNet

A complete integrated AI console that turns a goal into a matrix of routed work, assigns each node to the best AI, API, model, or deterministic tool, and returns a real output.

Chat with BoB. Let BoBNet route the work. Get the output.

Matrix-native execution Node-level routing Durable evidence

Launch position

BoBNet is being built for the work after the prompt.

The familiar chat interface is only the front door. Behind BoB, the console creates inspectable matrices, routes each node to the right intelligence, stores the work trail, pauses for approvals, and assembles final deliverables.

What changes

From one assistant to a net of specialized AI.

BoBNet is not one model trying to be every expert. It is an operating console for coordinated AI work, where each node can use the capability that fits.

01

Chat intake

Start naturally with BoB, clarify the goal, and convert intent into a runnable project shape.

02

Matrix planning

Break complex work into nodes with dependencies, assigned inputs, expected outputs, and launch readiness.

03

AI/API routing

Use the best model, API, local tool, validator, or reviewer for each specific node instead of one global model.

04

Output assembly

Return the finished artifact: files, reports, code, package records, research deliverables, or launch materials.

Node-by-node assignment

Every node can choose the right intelligence.

A matrix can mix frontier reasoning, local models, code models, OCR, search, browser automation, spreadsheet APIs, deterministic validators, and independent reviewers in one coordinated run.

Research nodeSearch API + long-context model
Code nodeCoding model + test runner
Document nodeOCR/API extraction + validator
Risk nodeReviewer model + approval gate
Publish nodeRenderer + artifact assembler

Why it matters

Complex work needs durable orchestration.

Leading AI tools have made conversation feel natural. BoBNet is built for the next problem: making the conversation produce a reliable, inspected, resumable output.

The BoBNet difference

  • Matrix graphs instead of loose prompt chains
  • Per-node AI/API/model/tool assignment
  • Durable state, evidence, and launch history
  • Operator approvals where risk requires it
  • Final artifacts instead of unfinished suggestions

White paper published

Read the BoBNet direction.

The paper explains why BoBNet is a step up from assistant-only systems and how the Matrix, Bob Hub, and node routing model work together.