Lehmus AI: Secure Generative AI Platform

Service status: Preview | Cost: Free (during Preview)

Lehmus AI is a private platform for high-performance Large Language Models (LLMs), hosted in the University of Oulu’s own environment. It provides a secure environment to experiment with AI and Retrieval-Augmented Generation (RAG) while ensuring your data stays within our local infrastructure.

Why use Lehmus AI?

  • Data Sovereignty: Your data never leaves the University of Oulu’s local computing platforms.
  • Secure Privacy: Fully compliant for handling Confidential and Personal data.
  • Seamless Integration: Use your favorite tools via OpenAI-compatible endpoints to connect directly to our hosted models.

What can Lehmus AI be used for?

  • Chat: Converse with language models. 
  • RAG: Use your own materials as a source of information for a language model. 
  • MCP: Connect language models to tools and data sources. 
  • APIs: Use language models in your own applications through an OpenAI-compatible interface. 
  • Preset Agents: Use preconfigured agents for common use cases. 
  • Guardrails: Guide and constrain language-model behavior with safeguards that support secure use. 

Access and Eligibility

  • Who can use it: Project rights are granted to staff members of the University of Oulu and Oamk. Students can gain access to the service through a staff member, such as the teacher in charge of a course.
  • How to get access: By filling the user request form (Requires University of Oulu or Oamk account).

Support and Learning

Maintenance and Reliability

  • SLA: 22 business hours for service and access requests. On Preview state, service may undergo unplanned service breaks and data losses may occur.
  • Standard Maintenance: Thursdays, 06:00 – 09:00.
  • System Upgrades: 3rd Tuesday of every month, 16:00 – 21:00.

Back to top