Analytics SaaS firm speeds service with Azure SQL Database serverless computing

Most businesses generate a lot of data. But that data isn’t valuable for decision making if it’s not structured and analyzed. And analytics aren’t as effective if they are slow and expensive to produce.

In 2017, Inlitix founders launched their company to offer a timely, easy-to-use, affordable BI as a service for small and medium-size businesses. To reach a broad spectrum of customers, they decided to focus on interoperability across different enterprise resource planning (ERP) systems with different datasets and offer out-of-the-box reporting and analytics solutions on top of this data. For each customer’s ERP system, Inlitix creates an environment entirely by APIs within Microsoft Azure, where Azure SQL Database serverless databases are provisioned and interoperate with Azure Data Factory, Azure Functions, and Microsoft Power BI. 

The company uses Azure Data Factory to pull raw data from a customer’s ERP system into staging tables in the serverless database before transforming it into a common data model inside the same serverless database. The common data model provides structure that Inlitix uses to provide high-quality BI across any number of customers at scale. It takes advantage of Power BI for reporting and analytics to gain insights across a variety of common business scenarios involving manufacturing, sales, purchase orders, inventory, and finance, among others. Inlitix keeps the reports and analytics current by triggering a refresh cycle using Azure Functions each time new ERP data is updated in the common data model.

However, Inlitix found that it only needs more compute power at certain times, such as when it spins up development and QA environments or when production data is refreshed. It didn’t make sense to pay for a constant high level of compute resources during the times when they weren’t needed. To minimize the company’s compute requirements and maintain affordability, staff would manually adjust and optimize queries, which took time away from focusing on more strategic tasks and adding value to their application. Inlitix wanted to optimize the balance between cost and performance of its databases while ensuring it could quickly process data from customer ERP systems and refresh Power BI reports more frequently.

The beauty of serverless computing

When Inlitix adopted SQL Database serverless, it could automatically scale the company’s compute capacity for its development, QA, and production workloads. “We provide software as a service and we like using platform as a service to do that,” says Alexandre Bouffard, Chief Technology Officer and Cofounder of Inlitix. “For us, Azure has always been outstanding compared to the competition, and we see a lot of value in what it helps us deliver.”

The company based its decision in large part on the seamless interoperation that it could achieve between SQL Database serverless and its existing environment, which includes Azure Data Factory and Azure Functions. Functions is a serverless compute service provided by Azure that allows developers to build serverless apps with database support. 

“We use Data Factory to push raw data into SQL Database serverless, where we transform the data to create well-structured, high-quality warehouse tables,” says Mathieu Viau, President and Cofounder of Inlitix. “Then we use Azure Functions to launch a Power BI refresh from the data in SQL Database. By combining Azure and Power BI, we’ve created an end-to-end environment that works together well, and it’s easy to monitor performance and manage security.”

Saved time, reduced costs

Before choosing SQL Database serverless, Inlitix had to balance a tradeoff between processing speed and cost because it paid a set price for its databases when using provisioned compute. Now, the company pays for only the compute it needs because SQL Database serverless automatically scales compute based on workload demand and bills only for the amount of compute used per second. “We can compute in bursts if we have a query that takes a lot of cores, and we pay for just a few seconds or minutes of compute instead of for continuous maximum capacity. By using SQL Database serverless to compute on demand, we reduced our development and QA environment costs by 10 times.”

Inlitix also increased efficiencies by reducing manual database management. When its serverless databases aren’t being used at all, they are configured to automatically pause during which time only storage-related usage is billed. When the team resumes development or triggers a data refresh in production, the database is automatically resumed. “We no longer spend a lot of engineering and development time optimizing our queries. We can maintain development and QA environments even if a project is paused and the environment just sits for a few weeks,” says Viau. 

The company is always looking for such savings to keep its solution as cost-effective as possible for its customers. “We’re keeping costs low while ensuring good performance by using SQL Database serverless. Depending on a customer’s usage pattern, the cost savings for production workloads can be as high as 15 times,” says Viau. 

Future-proofed services

Inlitix has noticed that the data needs of its customers change over time. For example, newer customers tend to have larger databases with millions of annual records, sometimes 20 times larger than Inlitix’s initial customers. “Our processing time could be fairly long for newer customers with considerably bigger databases, because we had to constantly optimize our queries to fit into a very small compute size database in order to get the price point we needed,” says Viau. “With SQL Database serverless, we don’t have to do that anymore.”

The company’s customers also tend to ask for more frequent reports as time goes on. Initially, new customers ask for refreshed data about once a week. After several weeks of using the data analytics to influence their decision making, they begin to request new reports once a day, then eventually prefer multiple refreshes per day. Concludes Bouffard, “Some customers want near real-time analytics because fresher data means more insights and better decisions. We’re confident that now we can easily meet changing customer needs by using SQL Database serverless.”

Find out more about Inlitix on Facebook, LinkedIn, X/Twitter, and YouTube.