Application developers rarely begin with a database feature. They begin with a problem.
How can I maintain live state for an AI application? How should I store and query JSON application data? Where should latency-sensitive payment authorization state live? How can I make telecom routing decisions quickly while maintaining active session state?
The Oracle TimesTen In-Memory Database samples repository now makes it easier to approach these questions from that perspective.
A new “Start here by use case” section in our github Oracle Samples Repository guides developers from a real application scenario directly to relevant sample code in Python, Node.js, and Java. Instead of first deciding which API, language directory, or database feature to investigate, developers can start with what they are trying to build.
Start with what you are building
The samples repository contains examples demonstrating how to use Oracle TimesTen In-Memory Database. With the new use-case-oriented entry point, those examples are organized around several practical application patterns.
AI and live application state
Modern AI applications often need more than model inference. They also need fast access to changing application state.
The TimesTen samples demonstrate patterns for response cache, chat memory, agent-workflow state, feature state, and other active application data. Developers can explore these patterns using Python and Node.js samples, with JDBC samples providing a Java path.
This provides a useful starting point for developers investigating where an in-memory relational database can fit into the operational data path of an AI-enabled application.
JSON application data
Applications increasingly combine relational processing with JSON documents. The new samples include focused examples for working with JSON from Python, Node.js, and Java.
They demonstrate JSON document storage, indexing, updates, and queries, providing concrete code that developers can examine and adapt when evaluating JSON-oriented application patterns with TimesTen.
Real-time financial authorization state
Payment authorization is a useful example of an application path where state must be available with low latency.
New Python, Node.js, and Java examples demonstrate TimesTen as a low-latency store for payment authorization decisions, idempotent replay, and hot risk state.
Rather than presenting these capabilities only as individual database operations, the samples place them in the context of a recognizable application use case.
Real-time telecom call routing state
Telecommunications systems provide another example where rapidly changing operational state can be central to application decisions.
The telecom samples show TimesTen as a low-latency store for routing decisions, idempotent replay, and hot session state, again with implementations available for Python, Node.js, and Java.
This gives developers a concrete scenario for exploring how TimesTen can support latency-sensitive state used in real-time decision paths.
Run the samples in containers
Exploring the code is only part of the development experience. The repository also provides deployment-oriented starting points. A modern demo container setup supports running the modern Python, Node.js, and Java samples with a local TimesTen container environment.
A note on TimesTen releases
For the modern demos covering AI and application state, payment authorization, telecom call routing, and agent workflow state, the repository requires TimesTen 26.1.1.1.0 or later.
Start exploring
If you have used TimesTen before, the new samples provide an opportunity to see familiar in-memory database capabilities applied to new application patterns. Visit the github Oracle TimesTen samples repository