Artificial Intelligence Takes on Large-Scale Database Management

Artificial Intelligence Artificial Intelligence (AI) is having a clear impact on databases. Among the major players, Microsoft is applying AI to continuously monitor Azure SQL database workload patterns and apply automatic tuning when it recognizes opportunities to improve the database performance, for example. And Oracle has unveiled the Oracle Autonomous Database Cloud, which uses Machine Learning to enable a database to automatically upgrade, patch, and tune itself while running.

Whatever database environments are at play in your enterprise, there is an opportunity to leverage AI as part of a Database Management solution to proactively support security, capacity planning, workload optimization, and performance tuning.

“We realized the requirements for handling large database environments were changing fast, and conventional tools simply weren’t keeping up,” said Rick Oppedisano, president and CEO of Delta Bravo. While monitoring is an absolutely essential foundation for Database Management, database security and predictive capacity planning are critical in managing a large environment, too.

“The cost of monitoring alone for these environments can be prohibitive, let alone trying to address the issues of database security, compliance, or predictive capacity planning – which conventional tools didn’t even address,” he said.

Oppedisano notes that what Microsoft and Oracle are doing – using AI and Machine Learning to ensure their products are configured and being used as designed – is important in reducing technical support calls for configuration-related issues.

“Where we differ is that we focus on each customer’s particular method of using those technologies beyond configuration. We help tune the database to best practices specific to their particular business requirements,” he noted.

Artificial Intelligence Digs In

Delta Bravo, Oppedisano says, came up with the idea of using AI, Machine Learning, Data Science, and Predictive Analytics to take on the full set of large Database Management issues after the company worked with AccuWeather’s Data Science team to build their D3 Advanced Weather Analytics platform.

D3 uses Machine Learning and AI to help companies quantify and predict the impact of weather on their business’s foot traffic, operational expenses, and product sales and to help customers adjust their operations in the most efficient way possible. The data for D3 comes from thousands of data sources including historic weather data, forecasted data, and its customers’ own data on sales, buying trends, and operational metrics.

D3 was a hit – but AccuWeather needed help managing the platform, including maintenance, security, and performance. Particularly with its customers’ data in the picture, high security assurances around the database were critical. D3 also runs on the Azure Cloud, and help was necessary for optimizing its performance and predicting the impact of future growth on cost.

Taking on the task, Delta Bravo came to the realization that it should build an AI Database Management solution to service the hugely complex AI system – one that could identify and prioritize issues and provide solution fixes at scale and with speed. Now, Delta Bravo has a couple of dozen clients leveraging the technology to predict issues for anywhere from hundreds to thousands of production and non-production databases.

“Our clients tend to be larger shops with smaller data teams – companies that have high expectations around security and performance of their data, but not a ton of people to deliver on them,” commented Oppedisano. These companies would like to put their teams in a position of saving time and being proactive instead of reactive.

Delta Bravo’s technology can determine, for example, that there is going to be a problem with a database’s performance 30 days before it happens in order for the company to minimize downtime and speed to resolution. Its Machine Learning algorithms help businesses understand the best pattern of use for a particular environment and then point out events that are creating anomalies, even before the thresholds that are set to generate an alert are reached. Knowing an issue is on the horizon “changes the way you are going to tackle solving the problem” – and for the better, Oppedisano noted.

“Delta Bravo forecasts issues and provides the code or step-by-step instructions for a fix,” he said. “It gives you all you need to know to manage a massive data tier.” It’s like having an expert doing the dirty work for you, then putting you in the ultimate position of making the choice of proceeding on those suggestions.

Its ability to give users a dashboard view into current trends and future predictions for resource utilization, performance, security, and compliance for each database helps in other ways, too. For instance, applying Predictive Analytics to database workloads can drive spend optimization in situations where clients may be considering migrating data to the Cloud but don’t have enough understanding of how it’s trending to know what that move will cost them six months down the road. “When you think of Data Management, you must keep an eye on growth trends, performance metrics, actions, and events,” Oppedisano said.

With all this information collected, AI can go into action, correlating past, present, and future database events; filtering out noise; and driving predictions of appropriate steps to take if, for instance, it finds a system spike tied to an event that negatively impacts query response.

“If it’s a performance problem, we run it through different models and find the solution with the highest likelihood of success. It’s the same with security and costs. Our models are retrained 10,000 times a day,” said Oppedisano.

The system can instantly secure databases to meet security and compliance best practices standards, too. Just as the technology can work its way through dozens of performance counters, each one weighted for a particular database’s use case, it assesses some 130 security counters (including abnormal query activity) to track key metrics associated with CPU or disk performance and predict their impact on the environment. It also lives up to the Department of Defense STIGS secure configuration standards, according to Oppedisano.

This is all important to companies like AccuWeather, which includes government agencies among its AccuWeather D3 customers. “We can show the violation, show the code our system used to find it, and the code we used to fix it,” he said. Big consulting companies can take months to accomplish the same ends: “Machine Learning and AI do it right here, generating the code and step-by-step instructions. This alone is a massive value. There’s no other use case where someone is using AI to do this.”

The Need for AI is Real

Another customer of Delta Bravo also showcases the need to bring AI into database operations support. The telecommunications company Comporium, according to Oppedisano, has thousands of databases that need to be tracked, secured, and maintained, but not a huge IT team, so using AI for greater efficiency is critical.

“Delta Bravo offers valuable, predictive insight on database security, capacity planning, and resource optimization,” said Andy Sember, IT Manager at Comporium.

“Delta Bravo organizes the information in a way that helps us understand immediate threats and what we need to tackle right away, as well as what we need to do to maintain higher ongoing standards. We see Delta Bravo as a key component in our ongoing data security initiatives.”

Comporium signed on with Delta Bravo for its security and optimization capabilities but expects downstream benefits in the DevOps space. When code is deployed, Delta Bravo details who made what change, what code was used, and both the current and predictive impact on the system. It also provides recommendations for how to remediate any issues. This creates visibility throughout the DevOps pipeline and enables developers to resolve 90 percent of database performance issues before going to DBAs, network or infrastructure teams.

“From the DevOps standpoint, a big thing is having everyone work together and promote code in a continuous fashion. If you have to go back on something you need to know happened, when and why, and what to do to fix it quickly,” Oppedisano noted.

By supplying CIO-level metrics – f or example, what the growth in SQL Server would cost the business in an aggregated way across thousands of servers – Delta Bravo’s solution also may help propel data teams to a seat at the executive table. “Dealing with databases now is about being a steward of data,” Oppedisano opined. “It’s about creating value for all the different audiences that interact with the database.”

Expect in the near future to see Delta Bravo update its security AI to include GDPR compliance and potential relationships with managed service providers who can bring its technology to their customers. It plans to release a Fed Cloud version of Delta Bravo before the end of Quarter 1.


Photo Credit:Dmitriy Rybin/

We use technologies such as cookies to understand how you use our site and to provide a better user experience. This includes personalizing content, using analytics and improving site operations. We may share your information about your use of our site with third parties in accordance with our Privacy Policy. You can change your cookie settings as described here at any time, but parts of our site may not function correctly without them. By continuing to use our site, you agree that we can save cookies on your device, unless you have disabled cookies.
I Accept