Conduct your Galaxy with a sentence.

mAesTRo is an MCP server for AVEVA System Platform. Connect it to Claude Desktop, Cursor, or any AI client that speaks the Model Context Protocol, and build, edit, audit, and deploy ArchestrA objects in plain English without leaving the tools you already use.

Join the Waitlist See How It Works

The Problem

AVEVA System Platform was built for engineers. The tooling around it wasn’t.

If you work in a Galaxy every day, you already know:

Repetitive object work eats your week.
Rolling out 200 templated instances, fixing UDAs across a hierarchy, normalizing scripts. It is all click by click, host by host, hour by hour.
Change is invisible until something breaks.
There is no clean way to see what changed in the Galaxy between Tuesday and Friday, let alone who did it or why a deploy started failing.
Knowledge lives in three engineers’ heads.
New hires learn the IDE the slow way. Tribal lookup tables and “ask Steve” don’t scale, and they walk out the door with the senior who built them.

You don’t need a new GUI. You need a faster way to run the platform you already trust.

Features

mAesTRo speaks Galaxy.
Not a general purpose chatbot pointed at a database. Every tool call maps to a real ArchestrA operation.

Natural Language Galaxy Operations

Ask in plain English: query objects, edit attributes, create instances, deploy templates. mAesTRo exposes these as MCP tools your AI client can call. Tool calls execute against your live Galaxy through AVEVA’s official COM interface (GRAccess). No UI scraping, no screen automation.

Change Management With Snapshots

Capture the Galaxy state before a change, run the change, and diff the result. Roll back to any snapshot. Every change-managed operation is logged.

Graphic Export And Import

Export ArchestrA Toolbox graphics for diffing, editing, or external version control, and import them back into a Galaxy.

Bulk And Batch Operations

Apply the same change across hundreds of objects in one pass. UDAs, extensions, scripts, choice groups, linked symbols. What used to be an afternoon is now a sentence.

Fast Search Across The Galaxy

Find every object using a given UDA. List every script that mentions a tag. Ask *”which objects still reference $LegacyMotor?”* and get an answer in seconds.

Galaxy Data Never Leaves Your Network

mAesTRo runs on your engineering host. The MCP server binds to localhost. Your AI client decides which LLM provider receives context, and you control what context it sees.

How It Works

You talk to your AI client.
Open your AI client (Claude Desktop, Cursor, Claude Code, etc.) on your engineering host and describe what you want, for example: “create 30 Pump instances from $Pump_Centrifugal under Area_05, set MotorHP to 25 on each, and stage them for deployment.”
mAesTRo plans and confirms.
Your AI client translates your intent into structured tool calls against the mAesTRo MCP server. mAesTRo previews exactly what is about to happen (objects, attributes, parents) and waits for your confirmation.
mAesTRo executes against your Galaxy.
mAesTRo runs the operations through AVEVA’s official COM interface (GRAccess), the same path the IDE uses. Results stream back into your AI client with object IDs, statuses, and any errors.
Everything is logged and reversible.
Snapshots, diffs, and audit logs are captured automatically so you can review, share, or undo any session.

For System Platform Engineers

Stop click farming through the IDE. Use mAesTRo to bulk edit attributes, refactor templates, validate UDAs across a hierarchy, or stage a coordinated deployment in plain English, with preview and snapshot based rollback. mAesTRo uses the same COM surface as the IDE, so anything mAesTRo does, you could have done by hand. Minutes instead of days.

For Engineering Managers

Get visibility into what your team is actually changing in the Galaxy, who changed it, and when. mAesTRo turns every change managed operation into a logged, diffable, reversible event, so code review patterns finally apply to System Platform work. Standardize how your team builds templates and rolls out instances. Onboard new engineers in days instead of months by letting them describe what they want and learn the IDE alongside the result. Quote and scope projects from real change history instead of guesswork.

For Plant Ops Leadership

Most unplanned downtime starts with a Galaxy change. mAesTRo adds a safety net: change managed operations are snapshotted before they run, previewable before they apply, and logged with a human readable diff. When something does go wrong, you can roll back to the snapshot with one command, no need to reconstruct what happened from memory.

See mAesTRo in Action

Bulk Per Instance Updates
An asset spreadsheet lands on your desk with serial numbers, locations, and motor sizes for 400 pump instances, each row a unique value for a different object. Hand the spreadsheet to your AI client, mAesTRo maps rows to objects, previews the writes, deploys. Hours of click work done in minutes.

Find What Changed
A script on your $TankLevel template started throwing errors in production after this week’s changes and nobody on the team remembers exactly what shifted. Ask mAesTRo for the diff between last week’s snapshot and the current Galaxy, isolate the edit, review it, fix or revert.

Search the Whole Galaxy
You need to retire the $LegacyMotor template but do not know which instances still derive from it. Ask mAesTRo. Get a clean list with parent paths and last touched dates. Refactor with confidence instead of grep and prayer.

Roll Back Under Pressure
It is 2 AM and a Galaxy change made last week is causing alarms to flood the control room. Pull the snapshot, diff against current state, roll back with one command, restore stable operations. Every action mAesTRo took is captured in the change log for the incident review.

Pricing

Pricing for individuals, teams, and air-gapped sites.

mAesTRo will be available in tiers for individual engineers, engineering teams, and multi-site enterprises. Planned tiers include an offline-licensed option for air-gapped environments.

Final pricing and tier details will be shared with the waitlist before public launch.

Join the Waitlist for Pricing

FAQs

Does mAesTRo send my Galaxy data to the cloud?

No. The mAesTRo MCP server runs locally on your engineering host and only accepts connections from localhost. The only data that ever leaves your machine is what you choose to put into a chat message with your AI client. You control which LLM provider receives that context. Fully offline deployments are on the roadmap.

Which versions of AVEVA System Platform does mAesTRo support?

mAesTRo targets AVEVA System Platform 2023 and later. The full compatibility matrix and any prerequisites are shared as part of waitlist onboarding.

Which LLM providers can I use?

mAesTRo speaks the Model Context Protocol, so any client with MCP support works (Claude Desktop, Cursor, Claude Code, and others). Bring your own API key (BYOK), point at a private cloud deployment, or run a self hosted model. The LLM never talks to your Galaxy directly. It only talks to mAesTRo through your client.

How does mAesTRo actually execute changes against the Galaxy?

Through AVEVA’s official COM interface, specifically GRAccess. mAesTRo never scrapes the UI and never modifies the underlying database directly. It operates under your Galaxy account’s existing permissions, so if your account cannot do something in the IDE, mAesTRo cannot do it either.

What happens if a Galaxy operation fails midway?

Change-managed sessions capture a snapshot before destructive actions and stream every command’s result back in real time. If something fails, you see exactly which object failed and why, and you can roll back to the snapshot with a single rollback command.

How do I install mAesTRo?

At launch, mAesTRo will install via a standard Windows installer that bundles the MCP server, the GRAccess runtime, and a registration helper that wires the server into your AI client of choice. No manual environment setup, no separate package installs.

Does mAesTRo work with redundant Galaxies?

Yes. mAesTRo connects to a GR node the same way the IDE does, so any redundancy configuration that works for your normal engineering workflow works for mAesTRo. Failover behavior follows the underlying System Platform configuration.

Is mAesTRo a replacement for the AVEVA IDE?

No. Think of mAesTRo as a senior engineering assistant that sits next to the IDE. It handles the repetitive, bulk, and audit-heavy work, so the IDE can stay focused on the deep configuration work that needs a human.

Can I try mAesTRo before launch?

The waitlist is the front door to early-access pilots. We’re onboarding a small number of teams to validate workflows in real production environments before broader release.

Ready to conduct your Galaxy?

mAesTRo is launching to the waitlist first. Sign up to get early-access invites, pricing details, and pilot opportunities before public launch.

Join the Waitlist