Challenge

wealthAPI initially relied on MySQL but faced challenges in scaling its platform to manage the high volume of user transaction data. Each transaction included a brief textual description, such as "car insurance" or "gym membership fee," which needed to be processed and enriched to offer actionable insights. wealthAPI’s legacy database struggled to handle this complexity and volume, leading to bottlenecks in processing and delays in delivering recommendations to users.

"We quickly outgrew MySQL. We needed a solution that could provide real-time insights regardless of data volume," said Wolfram Stacklies, CTO and co-founder of wealthAPI.

Solution

wealthAPI’s system transforms raw bank transactions into actionable insights by leveraging Astra DB’s vector similarity search capabilities. The process begins with data ingestion, where transactions are published to a message queue for asynchronous processing. Each transaction is then converted into a numeric vector using Astra DB’s vectorize feature, enabling efficient storage and lightning-fast similarity searches. These embeddings are clustered to identify recurring payment patterns, such as monthly music subscriptions or yearly health insurance. The architecture handles high data volumes seamlessly while maintaining strict data security to protect user anonymity and privacy.

wealthAPI also built a highly scalable contract detection solution powered by Astra DB’s vector database. This solution was deployed in under three months, demonstrating the speed and efficiency of Astra DB in production environments. Additionally, Astra DB enabled wealthAPI to reduce response times by an estimated 90%, ensuring near-instantaneous financial insights for users.

By leveraging Astra DB’s scalability, reliability, and advanced features like vector search, wealthAPI seamlessly integrated the database with its AI models. This enabled wealthAPI to focus on improving its recommendation algorithms and enhancing user experience instead of managing database infrastructure.

Impact

With Astra DB, wealthAPI transformed its ability to process and enrich transaction data at scale. The solution delivered:

• Real-time insights - Transactions were processed instantly, allowing users to receive timely, relevant recommendations.
• Improved scalability - Astra DB effortlessly handled the growing volume of transactional data without requiring additional operational overhead.
• Enhanced user experience - Users benefited from personalized recommendations that helped them save money and make informed financial decisions.
• Scalability at scale - wealthAPI estimates it grew its user base by 1,000% in 2024, and Astra DB seamlessly scaled with them.
• Faster processing - Response times improved by an estimated 90%, ensuring financial insights were delivered in real-time.
• Better performance - Query response times improved significantly, with Astra DB’s distributed architecture eliminating previous performance bottlenecks.

"Migrating from MySQL to Astra DB was a game-changer. We went from struggling with slow queries to significantly faster response times, even as our dataset grew exponentially. With Astra DB, response times have improved by an estimated 90%, ensuring financial insights are delivered in real-time. We can also process transactions at scale with minimal latency," said Stacklies.

wealthAPI’s partnership with DataStax highlights the power of a scalable, high-performance database in transforming financial services. Astra DB empowered wealthAPI to deliver a robust platform that improves users' financial decision-making by addressing the challenges of processing transaction data at scale.

"We don’t want our team to spend time on scalability issues—we want them to code and deliver. With DataStax, scalability just works," Stacklies added.