Chocolate Finance Sweetens Customer Experiences With Fresher Data on Snowflake
By replacing batch reporting, Chocolate Finance gets clearer customer insights, adjusts campaigns as they run and plans its roadmap with more confidence.
24hMaximum prior reporting lag, now reduced to near real time
Industry
Financial ServicesLocation
SingaporeStory highlights
- Reducing the wait for customer behavioral insights: Chocolate Finance cut its reporting lag drastically, from up to 24 hours to near real time, a big advantage for a digital-native fintech that operates 24/7. Data from its AWS-based stack flows into Snowflake through native S3 stages, Snowpipe and the Snowflake destination for Amazon Data Firehose, giving Chocolate Finance a more direct path from event-driven data to current reporting and decisions.
- Shifting from reacting to customer behavior to acting on it: Fresher data gives Chocolate Finance a more accurate view of customer behavior, helping the team anticipate customer needs. The team uses those insights to iterate on campaigns while they are running and plan the roadmap for future initiatives with more confidence.
- Giving every team access to data and AI: With relevant data consolidated in Snowflake, every Chocolate Finance team member can access insights and AI tools within Snowflake. More people across the company can now automate and build outside traditional technical groups.
Video Transcript
Chocolate Finance runs on AWS and thinks on Snowflake, and that’s how we make life richer.
My name is Thomas, and I’m the Chief Technology Officer at Chocolate Finance. The name is simple: everyone loves chocolate, and we want people to be happy about their money. We invest in short-term, investment-grade fixed income securities, but we’re not just trying to offer the best rates in the market. Our mission is to raise the bar on personal finance in general.
Our application is built mostly on AWS, and I did not necessarily want to fragment the ecosystem without good reason. Snowflake has a number of components that make integrating with AWS very ergonomic, such as native S3 stages and things like Snowpipe, which can be attached directly to buckets as a notification destination. We also utilize features like the native Snowflake destination for Amazon Data Firehose to stream our event-driven data directly into Snowflake. This is the perfect marriage between AWS, where we produce the information, and Snowflake, where we ingest it.
One of the main challenges we had before introducing a real-time data extraction platform was that our extraction was essentially batch-based, and even if it was relatively frequent, there were clear consequences to relying on data that isn’t fresh. With Snowflake, the reporting time lag reduced drastically from up to 24 hours down to real time. For a digital-native fintech that operates 24/7, that is extremely important. In business terms, that also meant being able to automate more processes without relying on heavier engineering effort, giving the ability to build to more people across the organization outside traditional technical groups.
Reaction became pre-emption because we had real-time information. We could monitor customer behavior much more accurately and use that to iterate on things like campaigns even as they were running, as well as plan the roadmap for future initiatives with a lot more confidence.
The biggest advantage has been rooted in the synergy of data and AI, empowering people who were previously constrained by technology. Every team member in our company is able to access insights on Snowflake because all the relevant data is consolidated there, and they have access to AI tools within Snowflake to help them navigate and query that data.
Snowflake blends the benefits of build versus buy really well. For us, it’s a platform we build on, yet we receive the advantages typically expected from an off-the-shelf platform—specifically, high-level features that make integration much more streamlined and ergonomic.
