dale lane

Looking back at my career so far, and what this could mean for what comes next…

I did a few different jobs across a few years – rotating through Development, Test, and Level 3 Service teams. I learned a lot, about middleware and the role of messaging in enterprise architectures, and more importantly about what software development at scale with high standards looks like.

Another Level 3 Service role, another chance to quickly understand and implement fixes in production enterprise software. But this time in a new product space (business process management), and on a new platform (z/OS mainframes).

A client-facing role – helping clients to understand and adopt emerging technologies. My projects spanned a wide range of technology areas, including IoT and sensors, web 2.0 and social, mobile apps, and AI and ML. This was sometimes through presentations and workshops, but often through building demos and proofs-of-concept.

In February 2011, the Watson research project was demonstrated on a US TV gameshow. For a year following this, the code was adopted by a small team of developers to take the research breakthrough, and commercialize it through working aspects like scalability, performance, security, and manageability.

The Watson platform derived from the Research breakthrough project was applied in large-scale projects in industries such as healthcare and financial services. I was responsible for the tooling used to develop and deliver the projects, with a focus on accuracy analysis for a large-scale complex machine learning system.

The Watson strategy shifted from a small number of large-scale dedicated clusters used by AI/ML experts on long-term specific projects, to providing reusable cloud services for developers without an AI/ML background to quickly create diverse projects. I worked on the creation of several new multi-tenant AI cloud services and the developer tooling to access them.

I worked on developing an idea for a new software offering (bringing the discipline of API management to event streaming protocols like MQ, MQTT, and Kafka) and prototyping the new technologies needed to enable it (leading to the creation of a first-of-a-kind Kafka-protocol aware Event Gateway).

Event Streams was Apache Kafka, with additional supporting components for running, managing, and monitoring it in Kubernetes clusters, and a focus on design and user experience.

We revived and released the “Managed Event Streams” prototype project that I’d worked on in 2017 (which we’d paused to build and launch Event Streams). This was a new type of product, enabling clients to use Kafka in new and different ways.

I worked with our clients to rapidly prototype and validate solutions using IBM technologies. In addition to building a wide range of Minimum Viable Products (MVPs) solutions for UK companies in a variety of industries, I also learned how IBM brings products to market and approaches sales, which I’d not seen working in product development.

I returned to Kafka, taking technical responsibility for the Event Streams and Event Endpoint Management products that I’d worked on before, and adding Event Processing to the set. Event Processing is an event stream processing platform based on Apache Flink, together with a business-user-friendly low-code event stream processing authoring tool.

And that brings us to now.

This isn’t meant to be a implied announcement that I have a plan for what comes next. (And I’ve written similar posts before in 2008, in 2013, and in 2021.)

But, I have been in my current role for a few years now, which my history suggests is a long time for me. It is a good time to reflect about what might come next.

I’d like to get back into AI/ML again – I miss my Watson days. Plus I watch more than a few AI projects today with frustration as they make many of the same mistakes that we made a decade or more ago.

I think there is more interesting work to do in the Kafka world – it’s not as widely adopted it could be. And Flink is only just getting started.

I’d like to go back to being more hands-on as a developer again – agentic coding tools are changing the game for the role of a technical leader. I’m excited about how this could improve the job of a “Chief Architect”.

But who knows what’s next? I’ve never been great at long-term career planning, and instead just looked for opportunities to build new things and solve new problems. Some of the most fun I’ve had came from jumping to something new outside of my comfort zone. ¯_(ツ)_/¯

Tags: career

This entry was posted on Sunday, April 26th, 2026 at 1:06 pm and is filed under ibm. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.