Product descriptions:
Kard, the first independent commerce media network, helps retail, restaurant and convenience brands capture new customers for life through hyper-personalized rewards. Analyzing over 16 billion transactions across their platform, Kard struggled as legacy systems slowed analytics and stretched resources. By adopting the Databricks Data Intelligence Platform, Kard was able to unify data engineering, analytics and governance, helping to streamline operations and reduce costs. With real-time insights at their fingertips, Kard now drives smarter customer growth and stronger loyalty.
Growing through legacy challenges
Kard helps retail, restaurant and convenience brands capture new customers for life through hyper-personalized rewards. The company connects merchants and card issuers, analyzing billions of transactions to deliver targeted offers and customer insights. As the business grew, Kard’s existing data systems struggled to keep up with the demands of large-scale analytics.
As a rapidly growing fintech startup, Kard dealt with increasing data volumes that pushed data pipelines to their limits — causing production jobs to stretch for weeks, ingestion pipelines to fail and manual analysis to slow insights to a crawl. “One of the things that I’ve run into working with large data and large data teams is that ingestion should be the simplest part, but it’s often the most painful and complicated,” explained Matt Schiros, Director of Data and Platform at Kard.
Beyond performance, Kard faced limitations in governance and collaboration that made it challenging for the lean startup to innovate and move quickly. Additionally, scaling was extremely costly and infrastructure management consumed engineering bandwidth and budget, which diverted focus from core business needs.
To reach their goals, Kard needed a platform that could scale efficiently, simplify operations and unify data engineering, analytics and governance. So they looked to Databricks.
A modern platform to compete at scale
Kard replaced fragmented systems with the Databricks Data Intelligence Platform, creating a single source of truth that scales effortlessly with growing data needs. Ingestion tasks that once took weeks now take hours, and Delta Lake with medallion architecture ensures reliable pipelines that support analytics and machine learning. Analysts can model billions of transactions in real time for segmentation, attribution and personalization.
Lakebase, a fully managed PostgreSQL OLTP engine, syncs billions of transactions from Delta tables, eliminating silos and manual engineering processes. Unity Catalog provides fine-grained access controls to establish the necessary governance and security, facilitating the safe management of sensitive transactional and PII data. For managing their machine learning (ML) models, MLflow helps to streamline model experimentation and versioning, while dbt empowers analysts to build models independently of engineering bottlenecks.
Integration with AWS has enabled Kard to realize faster insights, smarter targeting and the ability to operationalize data-driven experiences directly within their commerce media platform, helping retail and restaurant brands capture new customers for life. “The integrations are really straightforward and very easy to automate, providing the observability that we need around how those jobs are progressing. That has been a night and day difference from every other data platform I’ve worked with,” said Schiros.
By consolidating data tools into one platform, Kard now delivers faster, more reliable analytics while supporting growth across their partner ecosystem.
Innovating faster. Spending less.
With Databricks, Kard has dramatically improved speed and scalability. Ingesting four billion transactions from a partner that once took a month now takes just one engineer an hour. Fresh data can be activated almost immediately, accelerating campaign performance for merchants and issuers.
Serverless compute and autoscaling have dramatically reduced cloud spend by thousands of dollars a month while eliminating the need for dedicated infrastructure management effort. Compared to their legacy MongoDB infrastructure, Kard estimates the Databricks Platform is 10x more cost-effective at their current scale, which sets them up well as they continue to grow.
Governance and collaboration improved across the organization, giving analysts and engineers the freedom to explore data safely without DevOps support. This streamlined access fuels richer segmentation, more accurate forecasting and better customer personalization. “I think your strength is your platform. You just give people more tools to work with in the ecosystem,” added Schiros.
Kard now moves faster, operates leaner and uses data to fuel measurable growth. For a lean startup, that means faster insights, smarter targeting and the ability to operationalize data-driven experiences directly within their commerce media platform — helping retail and restaurant brands capture new customers for life.