Database administrators (DBAs) are under increasing pressure. The explosive growth of data, the rise of AI projects, and the fast-paced DevOps culture demand new ways of working. Developers are being given more responsibility, but often lack sufficient knowledge of databases. Redgate Software, which specializes in database development and monitoring, is adapting its platform to enable DBAs to keep pace with this changing reality. We discussed this with CTO Graham McMillan.
Before our conversation with McMillan, the CTO shared with us the results of a recent survey on the state of database management. Three main problems emerged. Security tops the list, followed by the increasing complexity of multi-database architectures. The third challenge is the growing volume of data, mainly caused by AI projects that generate enormous amounts of information.
According to McMillan, the database landscape is not evolving rapidly, but is shifting steadily. “Automation used to be a challenge, but now people expect it to be there,” he explains. Productivity today comes from automated processes, as this is now considered ‘good practice’. Nevertheless, the pressure on DBAs is only increasing. Problems are becoming more complex, and developers often lack in-depth knowledge of database structures, which places the burden on DBAs.
Collaboration between developers and DBAs
According to Redgate, the solution lies in closer collaboration between developers and DBAs. “Developers generally don’t have a good understanding of data and databases,” says McMillan. Tooling and monitoring can help bridge the gap, but the key lies in shared responsibility.
The concept of platform engineering plays an important role in this. By supporting developers with smart platforms and built-in security measures, developers can prevent errors and work faster simultaneously. “Developers don’t want to be slow,” says McMillan. “If you show them that certain working methods actually speed them up, you create a win-win situation.”
DevOps is living proof of this. It combines the roles of development and operations and forces teams to work together on stable, efficient deployments. Redgate’s internal philosophy is “you build it, you run it.” Teams that build software are also responsible if something goes wrong. “Whoever makes it keeps it running. That prevents people from making ill-considered choices,” says McMillan.
AI increases complexity
AI offers new possibilities, but also puts extra pressure on database management. Redgate is responding to this with SQL Prompt, with AI, a tool that works directly within SQL Server Management Studio. It helps users write code while retaining complete control. “It helps people write code, but they can still inspect everything,” says McMillan.
Nevertheless, he believes that AI is not yet ready to fully control databases. “No one is ready for that yet; it’s just not there yet.” It’s all about trust, understanding what AI is and isn’t suitable for.
DBAs value reliability and consistency, while AI models are inherently probabilistic. The models work by calculating probabilities. Redgate is therefore investigating applications where those functions are less critical. For example, AI can help narrow down problems: reducing the search area in complex situations. In Redgate’s monitoring product, for example, machine learning is used to identify patterns, such as monthly payroll peaks. “If it’s the third Thursday of the month, the system knows that’s normal. But if it happens a week earlier, it sounds the alarm.”
The future of the DBA remains human-centered
According to McMillan, the work of the DBA will not disappear, but change. “Everyone talks about jobs disappearing, but it’s more like widening a highway: as soon as you create more capacity, people fill it with new requests.” The emergence of low-code and no-code tools is also creating a new category of users: amateur developers who want access to production data without database knowledge. “The CFO building his own analytics app is really scary,” says McMillan. Yet it’s hard to say no to that.
That’s why soft skills are becoming increasingly important. The ability to understand perspectives and set priorities determines whether a DBA is successful. “It’s about identifying the real core problems and knowing which ones to solve first,” he explains.
Redgate also applies these principles internally. The organization uses benchmarks, such as the number of pull requests per team, to measure development speed but avoids imposing challenging targets. “Context is crucial,” says McMillan. “We want teams to learn from data, not be driven by it.”
Redgate’s path to Agentic AI
While most AI discussions revolve around technology, Redgate has opted for a practical, trust-based approach. The company is working on a phased introduction of Agentic AI within database management. In doing so, it chooses systems that can act autonomously, but always within controlled limits.
Instead of aiming for full autonomy right away, Redgate is developing pre-built templates: ready-made building blocks that enable small, safe steps. Think of a template that identifies a blocking query, temporarily stops the user, and sends log files to an analysis environment. “These are decisions that you can easily reverse,” explains McMillan. By gaining experience step by step, confidence in AI-driven processes grows.
Once users feel comfortable, they can combine templates into more complex workflows. This approach lowers the threshold for automation without relinquishing control. What’s more, templates provide context and structure, something that is often lacking in traditional AI chat interfaces.
According to Redgate, the technical capabilities are already in place. In theory, Agentic AI can independently detect problems, modify code, and test changes before they are rolled out to production. The key question, however, is: do DBAs trust AI enough to let that happen? “The technology is there, but the question is whether the DBA trusts us with all the steps that are necessary,” says McMillan. Sending data to LLMs also raises privacy and compliance issues. So far, Redgate has only tested the technology in scenarios that it can fully oversee. How AI will behave in unforeseen situations remains an open question.
Focus on user experience, not model building
Interestingly, Redgate has consciously decided not to invest in developing its own AI models. “There are thousands of models available. Building our own model is too expensive and unnecessary,” says McMillan. Instead, the company focuses on experimenting with existing models and improving the user experience.
That’s why Redgate isn’t recruiting an army of machine learning engineers. The company prefers to seek out user researchers, UX and UI specialists, and API engineers who understand how people work with AI. Redgate has in-house experts for specific machine learning components, but its core focus is on user-friendliness and secure integration with external models. “The success of AI in database management does not depend on the latest technology, but on how users want to work with it,” says McMillan.
Trust thanks to the SQL Server community
What sets Redgate apart is its strong connection to the SQL Server community. Redgate’s database platform initially grew by focusing on this area, but now supports many more databases. The user group is loyal, engaged, and willing to provide feedback. “Having such a strong community gives us valuable input and builds trust,” says McMillan. The community’s openness helps Redgate refine new features and test realistic scenarios.
This community-driven approach is at the heart of Redgate’s strategy. By involving users early on and translating their feedback into concrete product improvements, the company stays close to the practical realities. The result is technology that doesn’t remain abstract, but actually addresses the challenges faced by DBAs.
In the coming years, the role of the DBA will remain indispensable, but it will be filled differently. Automation and AI will take over routine tasks, while people focus on the strategic and human side of data. Redgate sees it as its mission to make that transition safe, manageable, and transparent. Where others focus primarily on technological innovation, Redgate opts for a people-centered evolution based on trust.
Tip: Redgate tackles multi-database challenge with Data Modeler