Skip to content

mAIstro

that work with your systems.

mAIstro lets you build agents that search, query your systems and decide. Each step has its own model and guardrails, and every run is logged.

Talk to an engineerSee the connectors

The building block

Agents are built from nodes

Each node is one step: search, call a model, query a system, transform, branch or loop. Nodes work the same in every deployment.

console › maistro › agent editor
The mAIstro editor showing a support-ticket agent: guardrail steps, two model calls, a knowledge base search, a call to another agent and a final reply.
A support-ticket triage agent in the editor.

How agents fit together

Nodes, agents and multi-agent flows

The same building block at every level, so an agent one team builds can become a step in another team's workflow.

  • The node

    One step with its own settings: which model, which source, which threshold. Build it in the visual editor or write it in NTL, NeuralSeek's template language. Both give you the same node.

  • The agent

    Nodes stacked into a flow or chained side by side, with conditions, loops and parallel steps. Parallel steps mean a run takes only as long as its slowest node.

  • The multi-agent

    Agents can call other agents, sharing variables or running separately. An orchestrator can also pick the right agent, or a plan of several, from your agent registry.

Runs you can inspect

Every run is logged

The Inspector shows every step: what was set, when and how. Governance shows how often each agent runs, response times, guardrail activations and the agents each run called.

Every step, inspected

console › governance › agent details
The Agent Details view under Governance for one agent: invocation and latency tiles across the top, and beneath them the invocation tree showing which child agents each run called.
Agent Details under Governance: latency percentiles per agent, and the invocation tree beneath.

After a run

Inspect, map and schedule your agents

Beside the editor, mAIstro shows what a run did, how agents call each other, the agent as NTL code, and when it runs on its own.

  • console › maistro › inspector
    The mAIstro Inspector with a model step open, showing its prompt, token counts, timing and output.
  • console › maistro › inspector
    The Inspector timeline, showing each step of a run on a time axis, including steps that ran in parallel.
  • console › maistro › visualizer
    The Visualizer tab: a graph of support agents, with intake agents feeding an orchestrator and a triage agent handing off to specialist agents.
  • console › maistro › ntl
    The NTL tab: the support-ticket agent as lines of NeuralSeek Template Language with syntax highlighting, the function library on the left and an empty Agent Output panel below.
  • console › maistro › scheduler
    The Scheduler tab: a table of schedules with agent, last run, next run in UTC, description and an Enabled switch per row, and a Select or Create a Schedule panel beneath.

Slide 1 of 5

01 / 05

Inspector: step by stepThe Inspector opens each step of a run: what was set, when, and how it was processed. A model step shows its prompt, cache setting, token counts, model time and output. Values masked by Remove PII stay masked here too.

What you control

What your team needs to take an agent from prototype to production data.

  • A model per step

    Choose which jobs each model does, override the model on any node, or pass a model ID on an API call. Use a fast model to classify and a stronger one to write the answer.

  • You decide when agents run

    An agent runs when called from the API, or on a schedule you set by minute, hour, day or month. You choose which agents run on their own and which wait for a request.

  • The same guardrails, mid-flow

    Prompt-injection protection, PII removal and the profanity filter are nodes you can place anywhere in a workflow. A retrieved document or a tool result gets screened the same way a question does.

  • Compare models side by side

    Select models, enter a question and compare their answers side by side, with token cost and speed recorded. It works for both Seek and mAIstro.

  • Adversarial testing per agent

    Point Red Team Testing at an agent and run it whenever you like. It reports the agent's risks, the test evidence and recommended fixes. Run it again after every change.

Common questions about mAIstro

What teams ask before they build their first agent.

  1. Is mAIstro a visual builder or code?

    Both. A flow built in the visual editor can be read and edited as NTL code, and back again. Work that would take glue code elsewhere is one node a reviewer can read.

  2. Do we start from an empty canvas?

    Only if you want to. Describe the agent in plain language and the auto-builder drafts it, or open a marketplace template and point it at your data. Every node stays visible and editable.

  3. What happens when we have hundreds of these?

    The run record is built into the platform, so no team has to set it up. Every run is logged, latency is broken down per agent and child call, and tokens and cost are tracked.

Start building with NeuralSeek today.

Sign up online and start with a two-week trial, or talk to a NeuralSeek expert about running it in your own cloud.

Start nowRequest a demo