EVREN MLOps platform opened to Turkish universities via YÖK

EVREN MLOps platform opened to Turkish universities via YÖK

In a letter dated 17 June 2026 (no. E-19924119-010.99-41068), the IT Department of Türkiye’s Council of Higher Education (YÖK) introduced the EVREN MLOps platform, developed by the Presidency of Defence Industries (SSB), to every university in the country. According to the letter, the platform brings data preparation, labelling, model training, testing, publishing and operational use together in one central structure and is open to academics and students.

The most striking part of the announcement is the access model: users sign in with e-Devlet, Türkiye’s national e-government login, and GPU usage is paid for with credits earned by uploading datasets or labelling data. This news piece summarises the letter, the technical details highlighted in university notices and what it means for enterprise software teams. Aksiyon Soft has no partnership with SSB or EVREN; this is independent industry news.

In short

• What: on 17 June 2026 YÖK introduced EVREN, SSB’s end-to-end MLOps platform, to all universities.
• Scope: the full model lifecycle, from data preparation to operational use, in one central structure.
• Access: identity verification through e-Devlet at evren.ssyz.org.tr.
• Credits: uploading and labelling datasets earns credits, which are spent on GPU time; a starter credit assigned via QR code is valid until 31 July 2026.
• Audience: academics and students, especially in AI, computer and software engineering, data science, electrical and electronics engineering and defence technologies.

What does the YÖK letter say?

According to the letter, which Hacettepe University’s Faculty of Engineering forwarded to its departments on 18 June 2026, YÖK presents the platform as part of efforts to strengthen the domestic defence industry ecosystem, build strategic capabilities in artificial intelligence and support academia-industry collaboration. The aim is for academics to use it in research and for undergraduate and graduate students to use it in their projects.

The second key theme is a contribution-based credit ecosystem. Users earn credits by uploading datasets, labelling data and similar contributions, and spend those credits on high-capacity GPU resources for model development. The goal is to tie access to expensive hardware to scientific and technical contribution rather than financial resources. The letter also notes that a starter credit assigned via its QR code can be used until 31 July 2026.

What does MLOps on a single platform mean?

MLOps is the practice of managing a machine learning model’s journey from data to production, much like DevOps does for software. In most teams that journey is spread across different tools: labelling in one, training on another server, the model registry in a separate service and inference on an entirely different stack. Every hand-off copies data, loses version information and makes it harder to trace which model was trained on which data.

The EVREN MLOps approach gathers these steps into one central structure. The inference side has been reachable since April 2026 through the official Python SDK on PyPI, so a vision model trained on the platform can be called from applications using the same credit balance.

What details did universities share?

Update: after the YÖK letter, departments published their own notices and some added technical details. According to the 23 June notice from Karadeniz Technical University’s Software Engineering Department, the platform offers 48 NVIDIA H200 GPUs for model training and 8 NVIDIA A6000 GPUs for inference, and it is open to all citizens of Türkiye through e-Devlet verification. Sakarya University’s Department of Labour Economics and Industrial Relations shared the same hardware figures along with a step-by-step getting-started guide.

Institution and unit	Date	Key information
YÖK IT Department	17.06.2026	Official letter to all universities; six-stage MLOps scope, credit model, e-Devlet login
Hacettepe University, Faculty of Engineering	18.06.2026	Forwarded the YÖK letter to its departments
KTÜ Software Engineering Department	23.06.2026	48× H200 for training, 8× A6000 for inference; open to all citizens
ESOGÜ Mathematics and Computer Science Department	06.07.2026	Summary of the YÖK text and the credit model
Sakarya University, Labour Economics Department	no date on page	Five-step starter guide; 1,000 starter credits for undergraduates via QR code

The steps in Sakarya’s guide also summarise the platform workflow: sign in with the e-Devlet button, follow projects, datasets and training runs on the dashboard, create shared workspaces with teammates, upload and label data, then launch training on the H200 cluster.

What does it mean for enterprise software teams?

At first glance the announcement is aimed at academia. But the students building final-year projects on the platform today will join enterprise AI teams within a few years, bringing habits such as data labelling, versioned models and credit-based GPU usage. In our assessment, three consequences stand out for organisations.

First, a domestic training and inference infrastructure becoming widespread creates a new option for projects with data residency requirements. Second, connecting models trained on the platform to business applications is still integration work: identity, quotas, error handling and audit logging have to be solved in the organisation’s own layer. Third, the credit-based model changes budget planning, turning a line item from invoices into contribution and balance tracking.

Readiness checklist for organisations

• Have datasets been classified under KVKK (Türkiye’s data protection law) for whether they may leave the organisation?
• Are labelling standards (class names, box rules) written down?
• Is there a record mapping model versions to their training data?
• In joint projects with academia, is data and model ownership defined in the contract?
• Is there a clear owner for the middleware that will connect trained models to applications?

EVREN series

This is the second instalment of our series following the EVREN platform. The series began with the release of EVREN SDK on PyPI; in September 2026 an LLM inference service with 11 models opened, and the national AI platform EVREN was launched broadly. We cover the technical route to connecting the platform to enterprise systems in our EVREN API integration guide.

How can Aksiyon Soft help?

Our Samsun-based team works remotely with organisations across Türkiye and makes planned on-site visits when needed. Through our API and integration service we build the middleware, data pipelines and monitoring layer that connect trained models to ERP, CRM or field applications. To manage multiple data sources from a single point, we recommend our API and data integration platform solution.

The process starts with a discovery session, followed by a narrowly scoped MVP, two-week sprint demos and hypercare plus SLA-backed support after go-live. Teams looking for a roadmap on LLMs and model routing can start with our article on the AI-ops layer and LLM router.

Frequently asked questions

Who can sign in to the EVREN MLOps platform?

The YÖK letter targets academics and students. According to KTÜ’s notice, the platform is open to all citizens of Türkiye through e-Devlet verification.

Is the platform paid?

Usage is credit-based. Credits are earned by uploading datasets, labelling and similar contributions and are spent on GPU time. There is also a starter credit assigned via QR code.

How long is the starter credit valid?

According to the YÖK letter, the starter credit assigned via QR code can be used until 31 July 2026. Sakarya’s notice states that the amount is 1,000 credits for undergraduates.

What hardware does it use?

According to university notices, 48 NVIDIA H200 GPUs are reserved for training and 8 NVIDIA A6000 GPUs for inference. The YÖK letter refers to high-capacity GPU resources.

Can companies use the platform too?

As of June 2026 the announcement is aimed at academia. For enterprise terms, follow the platform’s own announcements; we cover those developments in the later posts of this series.

Is Aksiyon Soft an EVREN partner?

No. Aksiyon Soft has no partnership with SSB or EVREN. This article is industry news based on public sources.

Let’s talk about your project

If you want to connect AI models to your existing systems or tidy up your data preparation process, write to us through our contact form and we will map out the scope together in the first call.

Sources

• YÖK IT Department — letter to universities on the EVREN platform, with Hacettepe Faculty of Engineering cover letter (17 Jun 2026)
• KTÜ Software Engineering Department — EVREN Platform (23 Jun 2026)
• ESOGÜ Mathematics and Computer Science Department — EVREN Platform (6 Jul 2026)
• Sakarya University Labour Economics Department — “EVREN” MLOps Platform (undated)