Dalus, SysML v2 and the real benefits of MBSE: Interview with Sebastian Völkl

Dalus is another company that wants to transform the product development market. I had the pleasure of talking to their co-founder Sebastian Völkl. The interview took place on April 7, 2025 and is embedded below, the article summarizes the most important points.

Dalus was accepted into the prestigious Y Combinator program and is now one of the most exciting new companies in systems engineering. In this interview, we talk about the following topics:

• Startup Dalus - Dalus was born out of the observation that existing MBSE tools were inefficient and that there was a gap in the market for better solutions.
• Target group - Dalus is aimed at small and medium-sized companies that want to use MBSE to promote innovation and not just compliance.
• Sweet spot - Ideal customers are teams that have been working document-based up to now and are looking for their first MBSE tool.
• SysML v2 - SysML v2 forms the basis for interoperability, but is implemented in a user-friendly way without overburdening the user.
• Avoid silos - Dalus relies on tools that involve the entire team instead of training just a few experts.
• Create added value - Dalus focuses on using AI support to migrate from documents to usable models within a few days.
• Case Study - Dalus linked a customer's scattered information in a model and thus enabled new, well-founded analyses.
• AI and SE - Sebastian Völkl sees great potential in the fact that AI will fundamentally change systems engineering in the future.
• Other tools - Sebastian rates the competition positively and sees dynamic development opportunities for Dalus and other start-ups.
• Integration - Dalus relies on open interfaces and develops numerous integrations to avoid vendor lock-in.
• The future - Dalus is experimenting with new AI-supported functions to make systems engineering simpler, faster and more efficient.

Why Dalus was founded

Sebastian had already launched two start-ups: CampusCrafta platform for students, and HackerBCIa neurotech start-up. He didn't actually want to found another start-up. The trigger for him to do so was his co-founder Eliot Khachiwho had several years of experience in aerospace at Aerojet Rocketdyne. Sebastian, who works with modern software tools, was appalled by the inefficiency of the tools used (e.g. Cameo).

After the team had familiarized themselves with the topic for several months and conducted around 150 interviews, they realized that:

• There is a pattern on which a business model can be based
• There is hardly anything in this niche at the moment

After this realization, the decision was made to get something going after all. Dalus was integrated into the Y Combinator program includedwhich they have just successfully completed.

Target group: Companies that really want MBSE

The bitter truth: many of the companies that use MBSE today don't actually want it. Traditional industries such as aerospace and defense often use MBSE primarily to meet compliance requirements. What Sebastian also sees are companies where individual employees know the tools and methods from research and impose them on the company. Incidentally, I like to refer to this behavior as Ivory tower modeling.

The bitter truth: many of the companies that use MBSE today don't actually want it.

Sebastian Völkl

For this reason, Dalus tends to find customers in industries where compliance pressure is lower and customers are interested in real benefits. He cited the energy sector, data centers and semiconductors as examples.

The "sweet spot" for Dalus are companies that have enough maturity in the development process to choose their first MBSE tool, i.e. that have worked with documents up to now. These companies generally value a tool that:

• Quick familiarization enables
• Can be used by the entire team
• Easy to integrate
• Can handle versions with confidence

As a start-up, Dalus cannot solve all problems at the same time, which is why the current focus is on small and medium-sized companies with 20 to 300 employees. Although larger companies are also part of the customer base, processes are much slower there and the tool compliance requirements are more complex and time-consuming.

SysML v2 primarily in the background

Dalus is based on SysML v2but does not want to overwhelm users. In concrete terms, this means that SysML provides the data model, but does not overwhelm users with hundreds of options in the UI. Instead, Sebastian sees SysML v2 as an enabler for interoperability: the model should be exportable as SysML v2 at any time.

In Dalus, SysML v2 is mainly an enabler for interoperability.

Sebastian Völkl

Sebastian emphasized how useful it is to be able to build on a solid standard. Apart from that, the timing is perfect.

At the same time, however, this also means that the user experience does not start with SysML, but with the customer benefit. As an example, Sebastian cited the way Dalus visualizes part attribution (in the sidebar, not as part of the part). Benefits for the user are more important than standard conformity. This also means that the user cannot use all SysML features via the interface.

To avoid being forced into this corset in the first place, Dalus has decided not to build on any open projects, such as SysON or the OMG pilot implementation.

Tools must not create silos

Sebastian criticizes the traditional approach of many tool manufacturers as no longer up to date.

The traditional introduction of MBSE tools he unintentionally sees as anti-organizational: Usually only a few experts are trained, which leads to silos and slows down effectiveness. This is precisely where Dalus comes in: The performance of MBSE should benefit the entire organization, not just a small group of experts.

This approach is of course directly related to how classic tools are structured. Therefore, a modern tool cannot simply consist of a canvas with hundreds of toolbars and model trees, as is the case with classic tools.

Not every organization is open to this way of thinking. Therefore, Dalus' target group does not consist of companies that have been practicing MBSE for decades, but is aimed at teams that are encountering MBSE for the first time. For this target group, the importance of SysML also fades into the background.

Create added value quickly

For Dalus' customers, MBSE is not an end in itself. The decisive factor is to generate tangible benefits quickly. That is why Dalus attaches great importance to rapid migration. This topic plays an important role for many customers when selecting tools.

It is important that every single person in the entire organization understands why [MBSE and Dalus] is important to us. You have to close this hurdle relatively quickly and then everyone is on board. If you deliver a result quickly, [tool implementation] doesn't have to be a top-down decision.

Sebastian Völkl

Here, Dalus relies on AI and now automates 80 to 90 percent of these steps. It is usually possible to go from documents to models within a few days, or one to two weeks at the latest. PDF extraction and document analysis are used for this.

Dalus often takes on this process itself as a service and supplies customers with a finished model. This is set to change in the long term. The aim is to provide customers with powerful migration tools that they can use to carry out the process themselves.

Case Study: Impact analysis of a data center

One example of the concrete added value of Dalus is a customer from the data center construction sector. Central information was scattered around: partly in Google Docs, in Python scripts and other sources. The content was barely linked to each other, which made it difficult to conduct well-founded analyses.

Dalus dismantled the Python-scripts into smaller, meaningful chunks and linked these to the business requirements. These "chunks" could then be migrated to Dalus. The result: a coherent system model that enabled completely new questions to be asked. For example:

• Does the adjustment of a certain set screw lead to more sales?
• If a tank is enlarged - are the requirements still met?

The migration only took around one and a half weeks. From this point onwards, the company was able to quickly create new versions and systematically evaluate scenarios.

AI and systems engineering

AI is already valuable for migration today, but the long-term potential is much greater. The company is based in San Francisco, where AI is currently more dominant than anywhere else. This inevitably raised the question of how AI would transform systems engineering. Nobody knows what is possible and what it will look like. The whole thing is a big experiment.

The question is not if, but when the two worlds of AI and systems engineering will meet.

Sebastian Völkl

Other tools

Of course, Sebastian is familiar with the competition, but sees significant differences in the details. Until recently, he saw Flow Engineering as a tool that focused primarily on requirements management.

Apart from that, tools from start-ups continue to develop in many rapid iterations. Who knows whether Flow Engineering (or Dalus) will not move in a completely different direction in the medium term?

On the contrary: many players make a lot of people think, and that is a positive thing. Especially at the moment, it's more "friendly competition" rather than fierce competition.

He also sees it as inevitable that Dalus will compete with the legacy tools. But here, too, the new start-ups differentiate themselves: firstly, because they don't simply copy features, and secondly, because the market is different. In the case of Dalus, it is primarily small, fast companies.

Of course, the large legacy companies will also have to look at how they can become faster at some point. He is already seeing interest and Dalus already has such customers in its portfolio, such as Israel Aerospace Industries.

We work with the largest aerospace/defense company in Israel, Israel Aerospace Industries, which employs 14,000 people.

Sebastian Völkl

Integration

Integration is crucial for customers because they do not want a vendor lock-in. Dalus therefore takes on the task of orchestration, among other things. Dalus already has around 100 integrations in the pipeline. Prioritization is based on customer requirements. What is actually needed is implemented first.

The integration list is based on existing tech stacks. For example, every company uses platforms for software development. Dalus has initially integrated GitLab as it is used by a customer. GitHub has been put on hold for the time being, but will also be integrated as soon as demand increases.

Data integration as well as imports and exports are rather static and hard work. In the medium term, Sebastian is planning a marketplace where customers can select the integrations they need themselves. The question of team integration and the handling of complex workflows, such as the integration of simulations and the comparison of these with requirements, will be more exciting.

The future

Dalus has a clear goal for the future. The existing gap is to be closed so that companies can switch more easily from document-based work to MBSE. The core function of the tool is in place and customers are using it.

However, the interaction between AI and MBSE is still highly experimental at the moment. Sebastian sees the need for intensive experimentation in this area. He assumes that Dalus will develop features over the next twelve months that possibly nobody needs. At the same time, there will be many innovations that you will look back on in a year's time and wonder how you could ever work without them.

I measure the value of a feature by how annoyed users would be if it were suddenly missing.

Sebastian Völkl

A few examples: Should the co-pilot be used to automatically generate requests? Technically this is possible, but it doesn't make much sense. On the other hand, should the AI automatically make changes that would otherwise take three hours of manual work? That's an excellent idea.

Today, Dalus can generate a system model that is 80 percent complete from existing documents in just a few minutes. The result is not perfect, but it creates a solid basis on which teams can build and offers customers enormous added value.

Many people still underestimate how strong and powerful AI actually is. As a systems engineer, I would think very carefully about how I use AI. The software engineers who didn't pay enough attention to the topic two years ago now have to catch up relatively quickly.

Systems engineers should be experimenting with MBSE and AI now to be prepared for what's coming in the next few years.

Sebastian Völkl

Interested in trying out Dalus?

Finally, Sebastian pointed out that interested parties can register for a seven-day free trial. This is usually preceded by an onboarding call. As many processes are currently still largely manual, Dalus makes sure that new customers are a good fit for the approach.

The pricing is transparent. Prices start from 99 US dollars per month and are published on the website. Compared to similar providers, Dalus is in the lower third.

Conclusion

With this interview, Dalus joins the ranks of the AI and MBSE tools presented in this blog. Many thanks to Sebastian Völkl.

Dalus focuses on fast results, strong integration and ease of use to make MBSE truly useful and not just a compliance obligation. The company shows how fresh impulses and a consistent focus on user-friendliness can change the tool market in the long term. I am excited to see what we will hear from Dalus in the coming months and years.