EveryMatrix, a renowned gambling aggregator and a top provider of industry-leading SaaS solutions for iGaming operators, was considering a complete revamp of their legacy systems to deliver top-quality data services for their range of EveryMatrix products.
An integral part of their iGaming solutions suite was a business intelligence reporting system built on the .NET and MS SQL technology stack. This tool helped gaming operators monitor player activity, generate invoices, calculate revenue, and adhere to gambling legislation.
However, steady annual growth in the number of players, games, and sports events put a heavy burden on this legacy system. Because of serious scalability bottlenecks, it could no longer smoothly process the increasing amounts of data coming from EveryMatrix solutions, customers’ CMSs, and payment operators. Besides, the system was difficult and costly to maintain. EveryMatrix needed a new architectural and technological approach to efficiently manage their 100 TB data warehouse system and ensure sufficient scalability for years to come.
As a trusted provider of advanced technology services, big data solutions, and machine learning solutions, Intellias set to establish powerful mechanisms to collect, aggregate, and process data from all EveryMatrix products and ensure data-driven insights for all end customers and partners.
As a long-time partner of EveryMatrix, Intellias has helped them with building a sportsbook platform, setting up a payment processing system, achieving PCI DSS compliance, and more. Over six years of collaboration, we have grown the development team for our client to 110 specialists, with several subteams working on different aspects of our client’s business. One of the projects Intellias has been working on is an advanced data services platform offering rich data streaming, processing, and reporting capabilities.
To offer real-time, historical, and predictive insights into business operations, the Intellias team built a new reporting solution using the latest business intelligence technologies. This solution incorporates:
- an in-memory message broker for receiving data feeds from various sources
- a dedicated data warehouse for storing pre-aggregated multidimensional data
- a customer-facing reporting interface for building, presenting, and visualizing business data
Data collection and validation
The intake message bus, which runs in the cloud, collects data feeds from different sources: EveryMatrix products, customer CMSs (user data), and payment operators. This data is then structured, validated, and made available for further processing. Built on the Apache Kafka distributed streaming platform, the message broker is fast, horizontally scalable, and fault tolerant thanks to data replication.
Data integration and processing
A set of free open-source solutions handle the integration of source data, merging it into a cluster of PostgreSQL servers. We use Apache Airflow for batch processing, Apache NiFi for streaming processing, and Confluent KSQL for Kafka streams. By choosing these services, we removed cost constraints and made the system more scalable.
The source data from the message broker is extracted, transformed, and loaded into the centralized warehouse and dimension tables.
Our own custom visualization solution compiles and presents data to customers in the form of reports, dashboards, graphics, and widgets. Widgets can be integrated directly into EveryMatrix products — Casino Engine, OddMatrix, MoneyMatrix, and PartnerMatrix — to display context-aware data to players and customers.
To support the constant growth of the EveryMatrix user base, which now numbers as many as 5 million, we have come up with several innovative solutions that help our client efficiently deal with the influx of data from different sources and provide customers with actionable business insights.
Cloud computing farm
Google BigQuery allows us to tap into Google’s powerful computing capabilities to ensure fast processing and analysis of massive data sets. Large volumes of data are split into small chunks stored in Google Cloud and can be easily processed within seconds by renting Google’s resources. By requesting the thousands of Google nodes necessary to compute small portions of data in a few seconds, we achieve fast query dispatching and data collection from multiple machines, resulting in fast calculation speeds.
Events and data from each product are brought into a single streaming platform with a data collection speed of 1 million messages per second. We use Apache Kafka, Kafka Connectors, and Apache NiFi for the events hub and ingest processing.
The platform allows data experts to analyze incoming messages on the fly with a delay of just several milliseconds and respond appropriately. If an event is classified under one of the rules configured in the system, it’s forwarded to the customer at their request. Data specialists can also act on a new event by sending an email or text message, making a request for fraud detection, or verifying if the user meets the requirements of the risk assessment machine learning model.
Our team developed an effective load management approach to increase capacity and handle high loads during peak traffic spikes. Our horizontally scalable solution allocates users to different servers, which can be added to the ring as needed. We use a consistent hash ring to assign users to nodes and a load balancer to distribute user traffic across available machines. In case of a load surge, we can increase the number of servers to withstand the load.
Prediction and recommendation engine
We built a machine learning recommendation system that includes a mechanism for data-based comparison of similar users (based on gender, age, place of birth, behavior, etc.) and their buying preferences for products and games.
An interactive recommendation model suggests new games to users based on statistics. The model ensures 90% precision in predicting the value of recommendations for gamers. Implementing this model resulted in a rise in user buying activity and a tangible increase in sales.
Additionally, we provided a fraud detection system, a model for predicting when users will leave the website, and an A/B testing system for calculating metrics.
Intellias has made a meaningful contribution to the transformation of EveryMatrix, guiding the company in a new data-driven direction through our engineering and consulting services. Our strong focus on data management and analytics has helped our client understand their customers’ needs and provide better services as well as enable customers to make more informed decisions and build sophisticated systems based on data.
Following our deep analytics approach, which included a thorough analysis of KPIs for each operator based on end user data, EveryMatrix was able to make efficient decisions on further improvements. This has resulted in the implementation of machine learning-based solutions that give EveryMatrix important insights into end users’ behavior. From identifying and recommending the most relevant games to eliminating illegitimate user activity, the systems we’ve delivered cover multiple aspects of the iGaming business.
The solution we’ve developed brings these advantages to our client:
- An always-on architecture and a single-source data warehouse that works with the entire EveryMatrix product family
- The ability to scale tenfold, enabling our client to handle millions or even billions of transactions without increasing the response time for generating reports and performing real-time calculations
- Improved report runtime performance through data pre-aggregation
- In-depth reporting with ad-hoc drill-down to the transaction data level