Mation: process automation. AI-native, legally compliant, and cost-effective. 

Marten Schoenherr, CEO AWA Network/Mation.work

Sebastian Schumann, programmable communications for Deutsche Telekom

Introduction

It’s been too long! Marten and Sebastian return to TADSummit to help the community understand Mation. and how Deutsche Telekom use it.

“Mation helps us to quickly and legally implement GDPR-compliant products on our own infrastructure.”
Sebastian, Senior Architect, Deutsche Telekom\

  • AI Native. Use any self-hosted (e.g., Ollama) or commercial (e.g., Vertex AI or OpenRouter) LLM. Works directly with all major AI providers (OpenAI, Anthropic, or Microsoft Azure). Integrate any Mation-supported app in your favorite LLM via our MCP server.
  • Legally Compliant. Open-source. DSGVO/GDPR compliant (General Data Protection Regulation). Mation can run on our customer’s own hardware. Even the Mation team does not have access to customer data.
  • Cost Effective. Transparent pricing model without hidden fees. Package prices including all services and functions. Unlimited users, integrations, flows, and executions across all pricing tiers. Flexible upgrading and downgrading options.

BOAT (Business Orchestration and Automation Technology)

Marten opens with a frank review of the mature BOAT (Business Orchestration and Automation Technology) landscape. There are many low-code / no-code options available. Node-RED was a popular open source option used in telecoms. Last year in our review of Elang RED, we reference Node-RED often.

Marten makes a strong statement that Node-RED was feature complete. So why build an open source alternative? Node-RED is not no-code, its popular with developers. Zapier was the most popular framework for automated workflows (called “Zaps”) to perform repetitive tasks without any coding. Through a US customer they discovered n8n. Similar to Zapier but with more control, offering both cloud and self-hosting options.

They discovered for telecom applications commercial no-code options were getting expensive, especially given the number of transaction, which in messaging applications can get into the billions of transactions.

Mation integrates with most of the CPaaS providers. <rant begins> CPaaS providers are not in my good books, they are making money off massive amounts of robocalling in the US, and worst of all the robocalls leave voicemail, up to 6 voice messages per day for me. Needless to say, CPaaS need stringent regulation with fines and prison sentences for the executives! Telcos need to take urgent action as CPaaS grubby money making is polluting the PSTN and bringing about its end sooner. <rant over>

Mation also integrates with the popular AI models. Check out Active Pieces for a Mation competitor in that space. Overall, Marten is proud that Mation is feature comparable with the the large automation vendors. Yet focused on telecoms.

Everyone has an open source MCP

Mation has their own MCP (Model Context Protocol). I referenced VCONIC TADHack, which is building on the VCON MCP server. Its a popular method of integration/exposure. I then asked about what automation platforms he comes across in his projects, Interestingly, Node-RED is not that common. As Mation is an open source project, there are many deployments across hundreds of organizations using Mation, many of which are transaction heavy, but beyond word of mouth they do not have the details of implementations.

I then asked about pricing, and discovered they’re moving to a pure open source model. I show the existing model below for the sake if history. I love Marten’s frankness on the money made with services, not licenses.

We mention Jambonz recently released a commercial model. And Marten shared a simple fact, Jambonz largest license is $2k, its just 2 man days. Selling services is much easier than selling licenses.

Screenshot

I then asked about the typical Mation customer. Generally its telcos, though they have customers in other industries, e.g. construction. However the application is generally telecoms related. The Mation community aspans from single developers to large corporations.

A close comparable company is is Active Pieces, which is funded, while Mation is self-funded.

We then moved onto the importance of AI coding to Mation. It’s a small team, that uses AI coding extensively. The SDD (Spec Driven Development) approach promoted by Jason Goecke is exactly their process. Interestingly they do discuss publishing the context files they use for Mation development to help the community build with AI coding.

I asked about consent and lifecycle management used in VCON. And they’ve done it! Though when it comes to compliance their stock answer is self-hosting.

How DT uses Mation

Finally, I move to Sebastian and ask why they adopted Mation and what are their use cases. Its their core business orchestration and automation platform. Not a customer saleable product. Rather an enabler for many of the services DT sells.

It’s a wholesale platform that covers the usual voice and messaging services, but also RBM (Rich Business Messaging) (or RCS), identity, and network APIs. Before Mation Sebastian would use Node-RED to demonstrate features to potential customers.

And that led to a number of problems for customers in building and maintaining what was demonstrated. Hence Mation provides the graphical tool that customers can use to craft and develop for their specific scenarios.

With Node RED scaling was an issue. Bring your own X was important to DT, and its customers. DT did not want to be an AI provider, rather enable customers to use their preferred model. Mation made that possible. Combined with the trust and regulation under which DT operates. I’ve covered previously the lack of honesty in CPaaS, we recently saw in the US the state Attorneys General calling out several CPaaS providers on the volume of robocalling they pass through to US citizens. Trust matters across all markets, and CPaaS lacks that.

Flexibility extends to front-end applications, there’s a forms feature, that enables attributes from multiple APIs to be composes into a new API or service. This can be done with pre-sales or sales engineers working with customers. Development projects are not required to set up a customer.

For generating billions of transactions, e.g. with VCON, Mation provides a scalable platform technically and commercially. Mation has achieved an important role in helping customers meet their needs that harness the trust, privacy (GDPR and common law), and the flexibility customers seek. It’s been easy to integrate Mation into DT’s portal with little effort. So workflow automation helps DT sell the answers customers seek.

Sebastian highlights the dichotomy made possible with the inclusion of MCP (Model Context Protocol). For SMS there is a traditional campaign manager where the workflow is deterministic, while a customer can use an AI to coordinate a number of functions to achieve the a similar goal. Its what the customer prefers.

Marten highlights with Mation it takes less that a day for any REST integration. Those integrations can be internal (provision / billing) or external systems, includes catalogs of APIs, and internal monitoring and alerting. So workflows can include Grafana (monitoring and observability) with alerting to the technical team that ratchets up alert from emails, to SMS, to calls. It’s the core of telecoms made possible thanks to Mation.

Thank you Marten and Sebastian , it was great to catch up.

One thought on “Mation: process automation. AI-native, legally compliant, and cost-effective. ”

Leave a Reply

Your email address will not be published. Required fields are marked *