Company Overview
Our client is a global MarTech company specializing in enhancing digital experiences for B2C companies. They offer a suite of products aimed at customer acquisition, engagement, and retention. As an AI/ML-powered marketing automation and customer engagement platform, our client helps businesses transform how marketing and product teams connect with consumers and prevent data loss.
Business Challenges
Our client was using Vertica, which is not a modern data warehouse platform and requires heavy maintenance. Scaling data in Vertica posed a huge challenge for our client. They incurred several downtime errors, late binding issues and problems while JSON parsing.
Additionally, the client also faced problems while sharing data and invested a lot of time in writing queries as they were using Vertica. They were storing 12 months of data on Vertica, which limited their data storage capacity. To facilitate the new data coming into the system, the client had to remove some historical data, which resulted in them unable to create accurate reports and dashboards. Vertica relies on AWS High Availability for redundancy, which incurred additional costs to our client. They wanted to know how they can reduce data latency and improve performance.
Kasmo’s Solutions
Our experts at Kasmo understood these significant challenges which our client faced and leveraged Snowflake to create an enhanced architecture and provided the following solutions:
- Schema Redesign: Kasmo migrated the Vertica architecture to Snowflake Multitenant Architecture which satisfied the business needs for future scalability and demands.
- Daily Binding: Kasmo solved the late binding issue (binding anonymous users with the users’ permanent ID) for the client, which was previously taking place only once a day. With our experts’ assistance, this issue was resolved and maintained hourly with Snowflake.
- Data Sharing: Kasmo enabled seamless data sharing to Snowflake account, even with no previously existing Snowflake accounts. Our experts enabled data sharing to customers who are:
-
- Snowflake Users
- Non-Snowflake Users
- Our experts also enabled our client to share data with customers through Snowflake Marketplace, which created a new revenue stream for them.
- Query Concurrency Test: Our experts at Kasmo customized the recent architecture to allow approximately 1000 users to work at the same time with enhanced performance, which was initially not possible with Vertica.
- Query Benchmarking: Our experts optimized the queries in Snowflake, which reduced query runtime by three times.
Results Achieved
With Kasmo’s expertise, the challenges which the clients faced were solved within three weeks. With an enhanced architecture, our client received these benefits:
- They could conduct hourly binding with almost 2.7 million rows completed in 1 minute and 9 seconds
- They could now ingest bulk data faster and prevent data loss
- The updated architecture ingested 22GB and compressed data in 5 mins and 20 seconds
- Low cost incurred while bulk data ingestion
- Reduced data latency and they could capture data as soon as it flowed in their system.
- With data sharing capabilities enhanced, the client could easily and securely share data to other customers in all regions.