Overview
SymphonyAI helps retailers and CPG manufacturers drive profitable revenue growth through AI-enabled decision-making. Its customers include 15 of the world’s 25 largest grocery retailers, thousands of retail brands, and hundreds of national and regional chains. SymphonyAI’s 700TB data platform continually ingests and analyzes terabytes of transaction and inventory data and delivers actionable insights to their customers.
New data platform requirements
Symphony RetailAI needed to find a data platform that could help it continue scale for customer growth yet deliver increasingly faster and more complex insights, providing not just regular reports but also supporting real-time and speech-enabled applications. At the time, SymphonyAI relied on a number of different data platforms, including Netezza and Amazon Redshift for batch analytics, SQL Server for transactional analytics, and 1010data for in-memory cubes and aggregates. However:
- Redshift costs were high, couldn’t process real-time data or provide interactive response, and cube build processes took up to 20 hours, making it hard to increase functionality without violating SLAs.
- 1010data could not scale effectively for large data sets and its custom programming language made development difficult.
- Netezza was already at full capacity, would soon be EOL’d, and was really only effective as a batch platform.
Yellowbrick: Simplification, consolidation, and scalability
The company evaluated Snowflake and Google BigQuery, but neither offered the predictable price/performance of Yellowbrick. For example, although one vendor promised on-demand scalability, Symphony found that it would either need to subject customers to unacceptable 12-15 second delays for real-time queries as instances in the cloud were scaled-up, or would need to pay four times as much as Yellowbrick to keep those instances running 24×7.
SymphonyAI moved all workloads from Netezza, Redshift, and 1010data to Yellowbrick, with results including:
- Faster, more accurate reports get richer insights to customers faster: When on Netezza, SymphonyAI had to limit reports that were run up to 5,000 times daily to 10% data sets. Today, on Yellowbrick, the company is running those reports with 100% data sets and delivering them twice as fast. Compared to AWS Redshift, cubes built on Yellowbrick completed 3X-5X faster.
- Private data clouds: Yellowbrick is deployed across the globe by SymphonyAI in different juristictions, allowing them to run workloads where their customers are and meet data residency requirements.
- Ease of migration. Compared to what would have been required on other platforms, the company’s move to Yellowbrick has been fast and easy, with few if any modifications needed.