Astra DB

We use Astra DB to store vectorized data from all of our company's unstructured files. This allows us to query these documents similar to how we query our structured SQL databases. One use case is the storage of vector data for all documents used in secure data rooms that typically contain hundreds of documents. It allows us to provide our investors with the ability to ask questions about the documents without opening each one. This saves them time and leads to a better customer experience.

• Astra DB has improved our customer experience by providing a new and innovative way to review documentation of our offerings.
• Astra DB brings the ChatGPT experience to our internal documents and system.

Azure Databricks

Microsoft Visual Studio Code, Azure App Service, Azure SQL Database

We migrated all of our 6 business critical self managed DB clusters to AstraDB. Main business problem we are solving with Astra DB is to utilize On Demand capacity pay per use model for our seasonal traffic /capacity patterns. Second problem we are addressing is seamless maintenance and upgrade feature that comes with Managed Service. Scope of our use case is business critical transactional DB's for high availability and low latency usecases

Astra DB is able to scale up and scale down during our peak holiday season without any prior preparatory changes

• Positive impact is through ROI - saving $250k/year with migration from self managed to Astra DB managed service
• Team members spending less time on maintenance, patching and upgrades
• Number of additional features such as Streaming, Vector Search are additional features available
• Auto scaling feature takes out stress of estimating and preparing for critical peak periods

Confluent

Core in our Legal work assistant application used by thousands of lawyers. Used to be able to work with large context which is crucial in legal work, consisting of factual information and legal sources that need different ways of handling,. Also lawyers need to see the different type of information in relation to create final deliverables that are grounded in the context information.

• Quite well proportioned cost and performance
• Possible to include much information at ok price
• Great partnership communication

Azure AI Bot Service

Microsoft Azure, Slack, Microsoft 365

Maester leverages Astra DB’s advanced vector search and retrieval capabilities to create an adaptive AI system that continuously refines its responses based on real-world user interactions. Maester also relies on Astra’s performance for semantic clustering. Maester can suggest relevant prompts and related analyses by grouping semantically similar user queries. This approach improves user experience and drives feature adoption, highlighting capabilities such as advanced forecasting and custom financial reporting.

• We leverage Astra DB to support our self-optimization functionality. Our product delivers up to 50% greater accuracy with minimal overhead—no fine-tuning of large models is required.

Pinecone

We use Astra DB as the primary data store for our Enrich and Engage products. Enrich is an AI-driven financial transaction enrichment product composed of several models that identify valuable information from data provided either by core banking or digital banking platforms used by financial institutions or retrieved using open banking connectivity. Engage is a set of tools that help financial institutions improve their digital channels with a focus on financial literacy, financial management and making better decisions based on enriched data.

• Moving from previous solution (a self-hosted Cassandra) enabled us to move towards less compex, less labour-intense and ultimately less expensive solution.
• Compared to the previous solution, we have been able to move from servicing clients that had dozens of thousands of end users to clients serving millions - operationally, that would have been very difficult to achieve without Astra, especially with the latencies and level of service required.

Apache Cassandra, Scylla and Redis Cloud

Google BigQuery, Looker, Google Gemini, MySQL, Perplexity