Startup Profile
OptiMigration

Introduce yourselves

Explain in three sentences what OptiMigration is! How long has OptiMigration been in existence?

OptiMigration is an AI-powered platform that automates document and completeness checks in skilled worker immigration procedures. It enables employers, recruitment agencies and public authorities to obtain a reliable assessment within minutes rather than weeks as to whether an application is complete, plausible and admissible – and to identify exactly what is missing. OptiMigration is a spin-off from TU Darmstadt and has been funded since February 2026 through the EXIST start-up grant; the team is based in Frankfurt am Main and Darmstadt.

How did the idea come about, and what challenges did we have to overcome?

The idea arose from real-world experience: whilst advising international skilled workers and students, the same pattern kept emerging time and again. Applications rarely fail because of a person’s qualifications, but rather because of formalities – a missing apostille stamp, an uncertified translation, or a discrepancy in names between a passport and a certificate. Our guiding principle here is: Germany does not have a talent gap; it has a procedural gap. Heterogeneous regulatory frameworks: requirements vary depending on the type of residence permit, country of origin, occupational group, federal state and, in some cases, the individual authority. This cannot be covered by a single checklist. The challenges lay less in identifying the problem than in managing it. The requirements vary considerably depending on the purpose of stay, country of origin, occupational group and region, meaning that the issue cannot be covered by a single checklist.
At the same time, the documents arrive in formats that traditional software cannot handle: mobile phone photos, scans of varying quality, foreign-language certificates, stamps and handwritten notes. Because we work with highly sensitive personal data, data protection was, for us, an architectural issue from the outset rather than an afterthought – we opted early on for processing to take place entirely within Europe. Finally, we had to draw a clear line: OptiMigration checks for completeness and plausibility, but does not replace either a legal assessment or an official decision.

How does Optimigration work, and how does AI support this project?

The process has been deliberately kept simple for users: a file is uploaded, and shortly afterwards the platform provides structured feedback on which requirements have been met, which have not, and what further information is required. Artificial intelligence takes care of the part that has traditionally taken people the most time – reading and understanding unstructured documents in various languages and formats, and extracting the information relevant to the procedure. What happens next is crucial for us. The actual assessment statement is not generated as a free-form model response, but by cross-referencing the extracted information against the stored procedural requirements. This ensures that every result is transparent and reproducible, and every statement can be traced back to its source in the original document. The final approval always rests with a human: the platform prepares the decision, it does not make it.

How exactly does AI-supported automation work in routine checks such as data verification or completeness checks?

It automates repetitive tasks, not discretionary judgements. During the completeness check, the platform compares the submitted documents against the requirements profile for the relevant procedure and does not simply return a blanket ‘incomplete’ status, but provides a specific list of missing items – that is, an indication of which document is still missing and in what form. During data verification, it checks information across multiple documents for consistency – such as spellings, dates and job titles – and highlights discrepancies before the application is submitted. In addition, formal criteria are checked, such as validity, deadlines and the availability of required certifications or translations. The result is a clear verification status for each requirement, each with a reference to the source, as well as a prioritised list of the next steps. A check that would otherwise require manual, experienced processing and take between 60 minutes and several hours is thus completed in a matter of minutes – and delivers a consistent, documented result for every case. The practical benefit lies less in the time saved than in avoiding the cycle of follow-up requests, which would otherwise quickly add up to weeks in the process.

What does ‘open-source AI platform’ mean in your context?

We’d like to clarify this point, as OptiMigration is not an open-source platform in the legal sense of the term. It is true, however, that we are building on an open, European AI stack – so to put it as a question: what does an open, European AI stack mean in your context? For us, this means three things. Firstly, openness rather than dependency: where it makes sense, we rely on open-source components and models with open weights, and our architecture is model-agnostic. We can swap out models without having to rebuild the product – in such a rapidly evolving field, this is a strategic decision. Secondly, data sovereignty: processing and hosting take place exclusively within the European Union; personal data is pseudonymised before model processing; and customer data is not used to train models. Thirdly, traceability: our verification logic is disclosed to customers and is auditable. Anyone working with us in a procedural context must be able to trace every verification statement back to the original document – without this transparency, the solution would not be usable in the public sector.

Let’s take a look into the future: where might the project be in five years’ time?

Our aim is for OptiMigration to become the preliminary screening layer for skilled worker procedures – the intermediary between employers, recruitment agencies and public authorities that ensures only complete and verifiable files enter the process. For public authorities, this means fewer requests for additional information and shorter processing times; for companies, it means planning certainty; and for the skilled workers themselves, it means a process whose duration can be predicted. On the way there, we envisage three main directions: integration with existing official specialist procedures, so that we do not remain a stand-alone solution; expansion from entry to related procedures such as professional recognition; and, in the medium term, roll-out to other European countries facing the same procedural gap. The aim is a European, sovereign solution to a European problem.

What milestones has your start-up achieved so far?

It all began with the EXIST start-up grant at TU Darmstadt, with Prof. Dr Peter Buxmann as our mentor; the first milestone report has already been accepted by the project organiser. At the same time, we were accepted into TU Darmstadt’s HIGHEST network and into AI Startup Rising at hessian.AI, and through the programme we were able to take part in the RAISE Summit in Paris in July 2026. The most important milestone for us is the move to live deployment: our first pilot has been running since May 2026, and a second pilot with a DAX-listed company is also in the pipeline. From a technical perspective, we have established a fully European, GDPR-compliant architecture with auditable verification logic during this period. We’ve also achieved two successes in the competitive arena: we were semi-finalists in the Hessian Start-up Award in the ‘University Start-up’ category and won the International Founders Award 2026 at the state level in Hesse – the national final will take place on 2 September 2026 in Frankfurt. Organisationally, the team has grown to three people, and preparations are underway to incorporate the company as a GmbH.

Is your solution already being used in practice? If so, where?

Yes, it is in live use in two pilot projects. The first has been running since May 2026 with an agency specialising in recruiting nursing staff from Latin America, where we review application files before they are submitted. The second is a DAX-listed company that uses our solution to recruit international skilled workers; we cannot disclose the name due to a confidentiality agreement. In addition, we are in concrete discussions with local authorities, immigration authorities and chambers of commerce and industry to further integrate the solution into the administrative side of the process.