Facts

MaestroHub, the short version.

One fact per line, each with the date it was true and where to read more. Last updated 19 October 2026.

What MaestroHub is

MaestroHub is the industrial context platform, in Gartner's Industrial Data Management category. It connects machines and business systems, gives every value one meaning in a governed namespace, and lets people and AI agents use that context inside permissions and an audit trail.
Oct 2026More: It runs on your own infrastructure: an edge box, a virtual machine or your Kubernetes cluster. It is not a hosted cloud service.
Oct 2026More: It is not a SCADA, an MES or a long-term historian. It reads from and writes to those systems.
Oct 2026More: MaestroHub is an independent company and is not related to UiPath Maestro or other products with a similar name.
Oct 2026: MaestroHub is one product with one licence per factory, plus a free trial. There are no editions.
Nov 2026More

Context Engine is live in release 3.0. It connects machines and systems, gives each value its schema, quality and origin, runs pipelines and buffers data when a link drops.
Oct 2026More: Release 3.1.0 shipped on 10 November 2026. It hardens store and forward for every eligible connector: one outcome per write, outages use no attempts and nothing is dropped silently. Upgrading from 3.0 runs upgrade apply once.
Nov 2026More: Release 3.0 shipped on 5 November 2026. It adds the lineage map, the approval gate for pipelines, delivery order per output and more.
Nov 2026More: The MCP server lets your own AI client, such as Claude, use MaestroHub. It is live in release 3.0.
Oct 2026More: The semantic layer is in the release of 5 November 2026, marked Beta. It gives every signal one meaning and one unit, and Find signals lists every signal with one meaning in one unit, as a table, a CSV file or one API call. Machines that look alike are suggested, and a person confirms each one.
5 Nov 2026More: App Studio and Maestro, the assistant inside MaestroHub, are in the 5 November 2026 release as Beta, both by early access on request. The Knowledge Graph is in development and not part of the release.
Nov 2026More

More than 100 connectors reach over 120 kinds of machines and systems, among them OPC UA, Siemens S7, Modbus, MQTT, EtherNet/IP, SAP, Ignition and the PI System.
26 Oct 2026More: 22 of the connectors can also write back, under permissions and audit.
4 Aug 2026: Every connector is included in every licence. There is no per-connector, per-tag or per-server charge.
Nov 2026More

Tier 1: one machine, single binary, 8 vCPU (4 physical cores with two threads each), 16 GB RAM, 200 GB SSD. One site. Context Engine runs with its embedded MQTT broker and embedded time-series database.
Oct 2026More: What one Tier 1 machine takes in with nothing lost is measured and published, dated, for 4, 8 and 22 vCPU. The published Tier 1 specification is the 8 vCPU measurement.
14 Oct 2026More: Tier 2: one machine running K3s, 16 vCPU, 24 GB RAM, 500 GB SSD. Manages many sites from one instance.
Oct 2026More: Tier 3: three identical Kubernetes nodes, 16 vCPU, 24 GB RAM and 500 GB SSD each, with application-tier high availability.
Oct 2026More: Runs on Windows, macOS and Linux, as Docker images, on K3s and on any conformant Kubernetes, for amd64 and arm64.
Oct 2026More: History is kept as an operational window that you set, the same for every licence: one day by default, and from release 3.0 each topic keeps 1, 7, 14 or 32 days. Long-term history goes to your own system of record.
Oct 2026More

Context Engine is priced per site, not per tag.
Oct 2026More: MaestroHub costs $16,000 per factory per year. A factory is one site; one licence covers every MaestroHub server there, with high availability and horizontal scalability included.
Nov 2026More: The free trial needs no licence key. It runs on an evaluation licence with every feature, including Advanced Access Control and Approvals, in two-hour sessions you can restart.
Nov 2026More: Evaluation on your own plant starts with a six-week pilot: one VM, no production writes.
Oct 2026More

MaestroHub's information security management is certified to ISO/IEC 27001.
Oct 2026More: Access is role-based, with single sign-on over OIDC, SAML and LDAP.
Oct 2026More: Every change lands in a hash-chained, tamper-evident audit trail with a verification call.
Oct 2026: MaestroHub publishes one dated benchmark: Tier 1 on Google Cloud C3 machines with 4, 8 and 22 vCPU, measured on 19 October 2026 with MaestroHub 3.0. A value counts only when it is authorized, schema-checked, stored and published live, and values are read back at the end. On every machine size the limit was the processor; above it devices are slowed and nothing acknowledged is dropped. The rates for each machine are on the benchmark page.
19 Oct 2026More

Customers include Ford Otosan, BSH, Akkim, Eczacıbaşı and Togg.
Oct 2026More: BSH: 15% operational efficiency, 10% more output, return on investment in under six months.
Oct 2026More: Ford Otosan: 95% accuracy predicting CNC tool anomalies, $22,000 saved per CNC machine a year, implemented in under a month.
Oct 2026More: Akkim: 15% better steam use, 8 tonnes of steam saved a day, return on investment in under a week.
Oct 2026More

Offices in İstanbul, Türkiye and Düsseldorf, Germany.
Oct 2026More: Founders: Yasir Tunçer (CEO), Barış Velioğlu (CTO) and Aras Yiğit Çakmak (Chief Delivery Officer).
Oct 2026More: Investors: İstanbul Portföy (lead), INC Innovation Center, Fark Labs and Cevher Group.
Oct 2026More: Thirteen releases shipped between February and August 2026.
Aug 2026: Audience Award at the WMF 2026 startup competition in Bologna.
Jun 2026More

Write to info@maestrohub.com and we will correct it. Every fact on this page carries the date it was true.