DevOps Amsterdam

After a break, we're bringing the DevOps Amsterdam community together again.

Join us on November 10 at ING Cedar in Amsterdam for an evening of food, drinks, conversations, and three talks exploring some of the challenges shaping DevOps and platform engineering today.

And there's another reason we're excited to get the community together again: we're working towards DevOpsDays Amsterdam 2027. We'll have more to share about that soon. 👀

For this meetup, we'll go from AI-assisted incident investigation, through 23 real-world ways of using an AI agent in a platform engineering team, to the lessons learned from organizations building their own internal developer platforms.

Whether you're working in DevOps, SRE, platform engineering, cloud, operations, software engineering, or just curious about where all of this is heading, we'd love to see you there.

Food and drinks will be provided, and there'll be plenty of time to catch up with old friends and meet new ones.

Agenda
17:30 – 18:30 | Doors open, food & drinks
18:30 – 18:35 | Welcome
18:35 – 19:05 | Building AIOps with an Ironman Mindset – Nikhil Gupta
19:05 – 19:35 | 23 Ways to Use an AI Agent in a Platform Engineering Team at RTL NL – Anna Kosek
19:35 – 20:05 | 7 Ways to Fail at Building a Platform – Michael Coté
20:05 – 21:00 | Drinks, networking & socializing
21:00 | End

Building AIOps with an Ironman Mindset by Nikhil Gupta

The Ironman triathlon is about endurance, disciplined preparation and controlled transitions. Flow and reliability are no different.
When an incident starts, responders need to understand customer impact, identify affected dependencies and decide what is safe to do next. But the evidence they need is often scattered across dashboards, deployment records, service catalogs, runbooks and other knowledge sources. Under pressure, responders have to reconstruct that operational context.
In this talk, Nikhil Gupta explores an AIOps design that continuously prepares this context and makes it available when it's needed.
Using a payment-service incident as an example, he'll explore investigations triggered by confirmed customer impact versus those triggered by an earlier anomaly. We'll see how an AI agent can choose read-only tools, collect evidence and propose verification steps while keeping the human responder responsible for operational decisions.

What you'll take away

• How to identify context gaps that force responders to reconnect evidence manually during incidents
• How to separate deterministic context preparation from probabilistic AI reasoning and human decisions
• How to bound agent investigations using read-only access, tool-call budgets, stop conditions and evidence trails
• Ways to evaluate AI-assisted investigations using diagnosis accuracy, responder rejection, time to first validated fact and lead time before confirmed impact

🤖 23 Ways to Use an AI Agent in a Platform Engineering Team at RTL NL by Anna Kosek

At RTL Netherlands, the team built Thanos, an AI agent that started life as a Slack chatbot and evolved into a platform engineering agent with more than 20 capabilities.

It reviews pull requests on Azure DevOps and GitHub, queries Grafana using natural language, generates architecture diagrams, manages work items and release notes, analyses build failures and supports automated incident root-cause analysis.

But the most interesting lesson isn't simply what the AI can do.
It's the human-in-the-loop patterns, identity-aware operations and trust model that made engineers actually want to use it.

In this fast-paced session, Anna Kosek will take us through 23 real-world use cases, complete with practical examples, honest failures and patterns you can take back to your own engineering teams.
No infinity stones required. Just a Kubernetes cluster and an open mind. 😉

About Anna Kosek
Engineering Manager Cloud & Data Platform at RTL Nederland.

🏗️ 7 Ways to Fail at Building a Platform by
Michael Coté

Building your own internal developer platform can look attractive. But when you build it, you also own it. Suddenly, you're in the business of developing and maintaining a product alongside your actual business.

Drawing on 10 years of platform engineering case studies from large organizations, Michael Coté will walk us through seven common risks of building your own platform:

Expansive scopes. Underestimating cost. Project mindsets. Death by customization. Skills. Security and compliance. Resume-driven development.

More importantly, he'll look at what organizations can do differently, including:

• Discovering the full scope of platform engineering across organizational silos
• Calculating costs and performance end-to-end
• Evaluating build-versus-buy decisions for internal developer platforms
• Creating a platform engineering strategy and staffing plan
• Applying product management thinking to platform engineering

Learn from organizations that have already encountered these problems so, hopefully, you don't have to repeat them yourself.

About Michael Coté
Michael Coté studies how large organizations improve the way they build software to run and grow their businesses. He's been an industry analyst at RedMonk and 451 Research, worked in corporate strategy and M&A, and was previously a programmer.
He's also the author of Changing Mindsets, Monolithic Transformation and The Business Bottleneck, and co-hosts podcasts including Software Defined Talk and Software Defined Interviews.