EventPay is a fast-growing Belgian startup that handles payment processing for events, bars and nightlife. Since they need to process a lot of transactions and are looking to continue their growth, EventPay wanted to improve their existing PostgreSQL data platform. In a half–day workshop, we lent our open-source database expertise and helped their technical lead find the right path forward. 

EventPay’s PostgreSQL setup, which they’d initially built for smaller-scale operations, was starting to show some growing pains under the pressure of their increasing scope. Managing all those separate databases for individual events, plus keeping the central system humming for analytics, was becoming increasingly complex.  

They were facing bottlenecks, and their growth was starting to suffer. They needed a fresh perspective and some expert guidance to optimize their database infrastructure and put them back on the fast track. 

Once they started looking for an expert, they knew who to call. EventPay’s technical lead, Ewout, knew Guy Gyles, our Senior DBA, from an earlier function. He was looking for a PostgreSQL powerhouse, and that, combined with our reputation for open-source solutions and our collaborative approach, made us the perfect fit.
 

To get a better feel for what EventPay needed exactly, we set up a workshop together with their CEO and technical lead. This was the perfect opportunity to get to know them better and what their plans were for the future.  

Throughout the day, we dove deep into their business and technical setup. Ewout gave us some great pointers into the challenges they were hoping to solve. We identified a few minor issues (like setting the correct time zone) and three key areas for improvement that would help them grow sustainably and securely:

1. Smart data organization with partitioning for scalability and speed 
2. Enhanced search capabilities with text search to speed up data retrieval 
3. Upgrading the security around their valuable payment data 

Let’s dive a bit deeper into the tips we shared for all three areas. Hopefully, you’ll learn from it as well! 

EventPay was dealing with a growing number of events, each generating its own data. To help them manage this data more efficiently, we focused on partitioning, a powerful PostgreSQL feature. If you’re not familiar with it, think of partitioning as organizing your database into neatly labeled filing cabinets instead of one giant, messy room. 

In this case, we suggested implementing RANGE partitioning for their tables related to device logs. This would streamline their many smaller event databases and make it easier to centralize everything for analysis. 

From what EventPay told us during the workshop, we realized that efficient search was another key area for improvement. To help them quickly find what they needed, we listed three different options: 

• For TEXT fields, we showed them how to use tsvector and tsquery data types and how to create indexes specifically designed for searching through text content. 
• For XML fields, we pointed them toward XPath queries, which helps to locate specific data. 
• For semi-structured data, we recommended using the JSONB data type for improved indexing and search. 

Security is always important, and even more so when you’re dealing with payment data! To make sure that every level of EventPay’s database infrastructure was covered, we suggested a layered approach that covers everything from system hardening to access control and encryption. Unfortunately, we cannot share any details here for security reasons. 
 

With our list of recommendations in hand, EventPay’s technical lead was able to optimize their data infrastructure so it can keep up with their ambitions. Of course, we’ll be available if they have any further questions or if they need a guiding hand!