ASC strategy

“Strategy” refers to a general plan or vision to achieve our goals (and what those goals are). Strategy can sometimes have a negative connotation by being too abstract (or at a level disconnected from what we actually do on the ground), but there is a value in knowing the issues of our team is currently facing and what we need to work on together to handle them.

This page contains various current and past strategy statements, to show how our goals have evolved over time.

2026 Autumn: AI related

AI hype is everywhere, with many overlapping teams and projects. Below is what we think we can focus on for the rest of the year.

• Usual RSE work (RD)

  • We focus on research projects and other projects below the “production” level. (Most RSEs) [note: This is mainly research projects, but we also support services when they come to us]

  • When something can serve a broad audience, we offer it to higher levels for productization. (EG)

  • We are not primarily responsible for third-party purchased AI services, but can give general guidance to those purchasing them and trying to use them if they come to us.

• Local AI (TP)

  • Promote 100% local AI to our users. [note: including teaching and documenting it, including aalto.fi pages such as https://www.aalto.fi/en/research-art/best-practices-for-ai-use-in-research-work (EG for this doc page)]

  • We will push llm-gateway and speech2text approved and documented. [note: RD, MH push ITS to get this done]

  • We will try to get sufficient hardware resources for local LLMs. [note: For example procurement or some Triton resources]

• AI competence, usability, and security practices

  • We will work on and document usable {tools, instructions} for agent-based coding (local & non-local). [note: This requires figuring out the latest practices and documenting on the generative-ai-tools page. It requires constant work to keep up to date and go beyond the hype.]

    • This includes making at least one usable recommendation for using agents securely.

  • We will, as much as possible, make sure aalto.fi documentation is up to date with our best practice. (ST, EG) [note: Coordinate with “AI in Aalto” http://aalto.fi and RES here https://www.aalto.fi/en/research-art/best-practices-for-ai-use-in-research-work]

  • Expand and market the CodeRefinery “Coding with AI” lesson. (EG) [note: Improve the lesson with latest info. Lesson at {NeIC AHM, CodeRefinery, groups that invite us to talk, EG arrange other Aalto presentations?}]

  • We will keep our GenAI roadmap page up to date with realistic options for researchers (all the usable options, including useful ones that are not approved). (YT) [note: We need some place as our “home base” for what we recommend which we can quickly update. This could be moved to aalto.fi sometime, but that can come later]. The sister page in Aalto.fi should be synced where practical. (EG)