.png)
The partnership offers an alternative for options overseas. It is an easy-to-use LLM Gateway endpoint, fully hosted in Finland, automatically GDPR-compliant, and powered exclusively by renewable energy. The solution is provided in LLM-as-a-Service model, making integration and migration as easy as possible.
“Our clients increasingly ask for secure, local AI capacity that does not compromise governance or expose sensitive data to external jurisdictions,” says Siili’s AI Principal Tommi Pajala, D.Sc. “Verda’s Nordic infrastructure gives us a reliable sovereign option, enabling companies to accelerate AI adoption with confidence.
A sovereign AI option for privacy-sensitive use cases
The collaboration is designed for sectors where data protection and operational continuity are non-negotiable: finance, healthcare, public services, and large enterprises modernizing legacy systems. The solution is especially designed with the sovereign AI angle in mind: all of the compute happens inside Finland.
Through Siilli’s LLM gateway hosted on Verda’s lightning-fast, enterprise-grade NVIDIA GPU infrastructure, organizations can utilize open-source models in a highly secure environment without investing resources in building their own infrastructure.
Through Siili’s LLM gateway, hosted on Verda’s enterprise-grade NVIDIA GPU infrastructure, organizations can run open-source or custom fine-tuned models in a highly secure environment. This gives them advanced AI capabilities without having to build or maintain their own infrastructure.
Key benefits include:
- Sovereign AI endpoints hosted in Finland
- Zero dependency on U.S. hyperscalers for core AI workloads
- Easy access to many open-source models in serverless fashion
- Automatically GDPR-compliant architecture
- Predictable and cost-efficient pricing
- LLM inference run on state-of-the-art NVIDIA GPUs
Accelerating innovation by reducing infrastructure overhead and model costs
While AI labs and US hyperscalers focus on licensing models with tens or hundreds of billions of parameters, many clients increasingly look for denser open-source models that can provide better return on investment.
“Not every use case requires the latest frontier model available to date,” says Anssi Harjunpää, VP of Sales and Partnerships at Verda. “Some teams need to validate the PoCs at the lowest cost possible. Others need full visibility and control over where their data goes.”
The partnership introduces an LLM-as-a-Service approach that lowers the threshold for experimentation, and in the end speeding time-to-market when developing proof of concepts to production. Clients can run AI pilots, prototypes, and internal tools with minimal upfront costs due to low-cost endpoint calls and flexible consumption-based pricing.
This enables:
- Faster and cheaper AI PoCs
- Easier integration into existing architectures
- Rapid scaling from tests to production
- On par model performance with data kept within your trust boundary
100% Renewable energy
AI infrastructure is power-intensive, but Siili and Verda aim to show that performance does not have to come at the expense of sustainability. All Verda operations in Finland run on 100% renewable energy, with local cooling conditions supporting exceptional energy efficiency. This enables innovation performance with a significantly lower environmental footprint.
-
Benefit for Verda: Access to a broader client base
Through Siili’s extensive customer network across the Nordics, Verda gains access to organizations that have previously lacked local alternatives for secure AI capacity.
Siili identified Verda as a strategic partner through its mission to bring clients the most suitable AI tools and infrastructure options, enhancing its existing NVIDIA partnership with a complementary sovereign AI offering.
