One of the most compelling use cases for AI in software engineering isn't generating code from scratch. It's helping experienced teams turn domain expertise into reusable solutions faster.
A great example comes from Novadoc ECM BV, a Netherlands-based consultancy working with IBM FileNet environments.
The team was dealing with a familiar challenge. Many FileNet configuration activities were still manual, repetitive, and carried operational risk. Novadoc already understood the problem deeply and had the expertise to improve the process. The missing piece wasn't knowledge. It was the engineering bandwidth needed to automate and productize the solution.
That's where IBM Bob entered the picture.
According to the case study, a developer started with a partially completed framework on a Friday and delivered a working application by Monday. A task that was expected to take roughly two weeks was condensed into a single weekend.
What I find particularly interesting isn't just the speed improvement. The bigger outcome was the creation of a reusable solution.
Instead of repeating the same manual effort for every customer engagement, Novadoc now has a foundation they can deploy, adapt, and extend across multiple projects. That's where the value compounds over time.
The story also highlights how IBM Bob contributed beyond code generation. The team used it to:
- Document a legacy codebase
- Improve architectural decisions
- Accelerate application development
- Create a reusable framework for future work
This reflects a broader trend I'm seeing with agentic development tools. The biggest gains often come from helping teams transform expertise into repeatable processes rather than simply writing code faster.
In many organizations, the real bottleneck isn't knowledge. It's turning that knowledge into scalable, reusable solutions.
And that's exactly where this story becomes interesting.
📖 Read the full story: Novadoc ECM BV Case Study
Top comments (0)