Our CFO initiated data analytics training three years ago with a small cohort of finance professionals to promote citizen development.  New classes were conducted each of the last two years to employees representing a broader range of business domains. Our central data and analytics team has not kept pace with the growth of citizen development, and many of the citizen-developed reports use poorly designed models instead of curated models. We are looking to potentially create a new role around self-service governance.  For those who have created a similar role, do you have any learnings or recommendations?

Consider adding a technical leadership (governance) role to provide guidance on the design, as Marco mentioned. Establishing design principles and guardrails, such as promoting reusability, cloud-first solutions, and data as a product, will be critical in helping citizen developers learn the essentials when designing a report or model.

For training, consider using vendor computer-based trainings or recording instructor-led sessions for self-serve onboarding of newcomers, instead of relying on individual instructors.

One of the biggest challenges we encountered with low-code platforms and operators lacking a strong technical background is the absence of proper change management. Most SaaS solutions allow for rapid modifications without integrating a structured CI/CD workflow, making release management either extremely difficult or unmanageable. This issue is further compounded when merging functionalities, as low-code platforms—typically based on DAGs—offer little to no version comparison capabilities.

Regardless of whether low-code is involved, when building a system composed of multiple components, it is essential to have a principal engineer or solution architect guiding the design.

Ultimately, citizen developers are still developers implementing their own ideas, which can be risky without proper oversight. The real problem isn’t the tool itself, but rather the way the solution evolves without a solid architectural foundation or adherence to established architecture decision records. What you likely need is an engineering principal or solution architect to ensure the system’s integrity and maintainability.

Hi, this is a very common scenario across many organisations. A few points to consider, before you hire someone for self-service governance role:
- How well the curated models support the needs of the business? Is the business trying to fill a gap by doing their own model development? 
- Is the data and analytics team, along with data governance enabled to provide guard rails, as well as educate the business on new development. Depending upon your current org. structure, this team should be able to control or atleast influence what goes into production. 

Focus on increasing the adoption of curated models, perhaps data and analytics needs to engage better with business in understanding their needs. That's my 2 cents.

If you do get a new resource, it would make sense to get someone good at data modelling. Hope this helps :)