This week, we announced that we have officially rolled out mondayDB 1.0 to 100% of our customers. To mark the occasion, we thought it was only fitting to take you, reader, through the journey of just how the mondayDB came to be.
mondayDB is the new data infrastructure behind our monday.com Work Operating System, designed to drive your best work. Built by our monday.com team, the new data architecture uniquely suits the platform’s needs. It was created to adapt to the most extensive, complex work scenarios imaginable, allowing organizations of any size to build and manage workflows at scale while enjoying exceptional performance.
mondayDB 1.0, the released foundational version, allows loading boards with thousands of items up to 5X faster, enabling customers to work with data-intensive and complex workflows. It also improves the performance, user experience, and responsiveness of the monday.com Work OS platform, ensuring that massive Boards load in seconds and that key Board interactions happen nearly instantaneously.
1.0 is just the start. Future releases are planned throughout the coming years, providing additional speed improvements, scale, and extended functionality. However, let’s take a step back and start from the beginning.
The Origin Story
In 2014, when we introduced monday.com to the world (our company was called “daPulse”), boards were uncomplicated, and many of our customers were SMBs trying to solve everyday problems. For example, managing a few projects, each with a few tasks.
Tables (now known as Boards) were smaller-scale, and the reporting and aggregation functionality required for handling them, such as sorting and filtering, was effortless. There were way fewer column types, widget types, automations, and formulas. We addressed most use cases by providing a primarily client-based solution for showing and analyzing data and using a monolithic database for long-term storage.
Simpler times, with smaller Boards that easily fit in customers’ browsers
Catalyst For Change
As time passed, our customers grew, and our product needed to grow with them. As the years went by and our customer base diversified, the scale of Boards and Dashboards we hosted for our customers kept growing. The number and size of accounts we onboarded every week and the complexity and variety of data query use cases we handle are still growing. We were blown away by the incredible use cases and work processes our customers were using monday.com for.
We realized that the current infrastructure wasn’t allowing us to give customers the experience and performance they have come to rely on from the monday.com Work OS platform. These challenges led us to move to a new data infrastructure that had to solve two critical issues. It would have to be an immediate solution to our current problems, and it would have to be flexible enough to meet the required workloads as we keep scaling up.
Things are scaling up. More customers, more use cases, bigger and more complex Boards and Dashboards. Gotta go server-side.
Bringing it to Life
This idea for a new data infrastructure wasn’t a new one for us. We began toying with the idea a few years before we started building it. We knew this wasn’t something we could just dive right into, so we had to take an approach as strategic as possible.
We decided to begin with an initial evaluation of some existing off-the-shelf solutions. We would have scaled up by industry standards with MySQL, elasticSearch, or Clickhouse solutions. Those mentioned above are powerful tools for solving specific problems – providing real-time updates, handling a large data throughput, providing increased flexibility, or the ability to scale up indefinitely. However, the complexity of monday.com’s platform coupled with the scale and flexibility we aspire to, we weren’t willing to compromise on any of those requirements. We eventually realized we would have to resolve things internally and build our own solution.
After another year of exciting research and development, we entered the uncharted territory of technical challenges and creative solutions on a quest to create what is now mondayDB. This resulted in a robust data engine containing all the platform’s data, with the ability to query it quickly – specifically designed for the unique needs of monday.com.
By early 2023, the data infrastructure was ready to receive its first “internal client,” – and we started working on connecting our Boards using a newly built Query layer. The fruit of this integration went through successful Alpha and Beta stages, and last May, we began gradually rolling it out to selected customers.
A company-owned data management system has many benefits, the most important being that we can tailor mondayDB to the platform’s unique needs. For example, we can build custom solutions to focus on specific monday.com use cases and optimize them to our needs. This includes links between tables and aggregations on custom UDFs (formulas). We can also run the Resource Control – by separating storage from compute, we can scale horizontally and optimize our server usage according to a customer’s needs. Our ability to monitor and analyze our usage allows us to optimize our resources accordingly.
As mentioned, mondayDB will enable more complex workflows, bigger boards and dashboards, and more robust developer capabilities. While developing the new infrastructure, we were powered by the desire to increase the value of monday.com for our customers. mondayDB offers many benefits, but we want to spotlight the most impactful that will allow our customers to benefit holistically.
mondayDB in the Wild
While mondayDB is new to many of our customers today, we were lucky enough to test the updated infrastructure with select customers beforehand. We’ve already seen how impactful the switch to mondayDB has been for them and can’t wait for the rest of our customers to start experiencing these benefits.
Here are some incredible examples of the benefits mondayDB is already providing:
- A large semiconductor chip manufacturer uses monday.com as their primary work OS. That means their engineering managers and project management officers assign tasks using monday.com, and track the timeline in which their engineers perform design tasks. Customer Success Managers and Account Managers also use it to track customer requests and issues to ensure they are assigned to the right teams and resolved promptly. Since they use monday.com to manage massive, complex projects, they have Boards with up to 20K items. Before mondayDB was released, loading such boards took a significant amount of time, which hurt their efficiency. They now report that the same Boards load in up to 5 seconds, which is a considerable boost.
- An American multinational chain of fast food restaurants uses the monday Work OS platform to manage their multitude of restaurants, which means handling large volumes of employees, customers, providers, invoices, and receipts. In many cases, the data required to manage their restaurants have to be imported in bulk from an external source, such as Excel spreadsheets with thousands of records at a time. Before mondayDB was rolled out, such imports would take minutes or hours at a time, and in extreme cases – it would fail altogether. Once mondayDB had been deployed, the users were able to import massive item lists, have their automated workflows successfully trigger off of each imported item, and they were able to view the resulting Boards in under 4 seconds.
- A digital marketing agency uses monday.com as their Customer Relationship Management platform, otherwise known as CRM. That means that monday.com has to handle tens of thousands of new leads, contacts, deals, and customers every month. When viewing Boards with 20k leads, they encountered a poor user experience caused by long loading times. Thanks to mondayDB, they noticed that these Boards load up to 5X times faster and that the user experience is significantly smoother.
We’re thrilled to share the first phase of mondayDB with our customers. We can’t wait to see how they utilize it to power their most significant, complex work scenarios enabling them to build and manage workflows at scale without being limited by performance constraints.
This is just the beginning. We’re already busy working on the second phase of mondayDB, which we’ll begin rolling out to customers later this year, enabling Dashboards to load up to 10X faster. Beyond that, by the beginning of 2024, mondayDB will provide increased scale for the entirety of the Work OS platform, including automations, integrations, apps, and extending the monday.com open API.
If you’d like an even deeper look into the technical details of mondayDB, check out this post, drafted by Liran Brimer, Tech Lead, on our Engineering team.