AI Platform Engineering
Organizations are increasingly investing efforts in unlocking the value of Generative AI. It all starts with rallying organizations around Gen AI and developing ideas to bring more value to their customers and move the needle toward business goals. Ideation events (or “hackathons”) are a great way to engage the organization, and you can see how it worked for us and our clients in my DevOps Enterprise Summit talk.
But once the Generative AI initiative gets traction, the pattern is the same as with any other technology: the organization tries something, and that something works (it produces value). The organization would like to have more of it, so it needs to scale. The question is how to efficiently scale the approach, the technology, and ways of working across the organization.
Platform Engineering is the pattern that successful organizations use. Sometimes, people have this shortsighted view that platform engineering is only about building infrastructure and pipelines to deploy apps to Kubernetes. Platform engineering is about building abstractions over infrastructure, tools, and processes that help teams deliver value more efficiently. Yesterday, it was about Kubernetes. Today, it is also about providing access to approved LLM models, RAG mechanisms, vector databases, caching, model versioning, and connectors to legacy data sources, hence AI Platform Engineering.
Platform teams shouldn’t be static; we shouldn’t rely on them just to install Kubernetes and build application deployment pipelines. New paradigms are coming, and we need to find ways to support them through platform teams. Therefore, considering the speed of technology these days, platform teams should be on alert and open to supporting teams in new areas, with Generative AI being first in the line.
Interview of the Month
AI Platform Engineering with Patrick Debois
You know how in every good movie, it takes some time for a plot to develops, and the main character is introduced only later in the movie… We’ll that’s why it took me 43 episodes of the DevOps podcast to welcome a person that started it all! I talked with Patrick Debois about challenges of introducing Generative AI and how can AI Platform Engineering help.
Handpicked articles
The Death of the Junior Developer – Steve Yegge shares his view on what the future could look like for junior developers given the latest LLM developments.
Abilene paradox – I learned about this fallacy while reading Flow Engineering, and I haven’t been able to unsee it ever since.
Stanford AI Index 2024 Report – The report states that “the number of AI regulations in the USA has increased 56.3% in the last year.” US regulations are catching up with Europe, which correlates with organizations investing in Governance Engineering.
Read with us
Crucial Conversations: Tools for Talking When Stakes are High
Get the bookOnce, I did a poll on social networks asking people whether the main reason their last project failed was technical or organizational. Guess what, 92% blamed organizational reasons.
Crucial Conversations teaches us how to establish psychological safety and lead meaningful conversations that foster better understanding and conflict resolution.
Quote of the day
“Statisticians, like artists, have the bad habit of falling in love with their models.”
―George E. P. Box, British statistician
Sharing is caring
If you find 0800-DEVOPS useful, share it with your friends, check out the archive and subscribe to receive it in your mailbox.
