Why Your Cloud Architecture Can Scale Users But Not Costs

If you've built software for a while, you've probably seen a similar architecture diagram: database for transactions/relational data, Redis for caching, Kafka for events, OpenSearch for full-text and keyword search, object storage as a data lake, and maybe a vector database for AI integration. There's nothing wrong with these choices: each technology is good at its job. For example, you'll find countless articles, papers, and engineering blogs extolling the virtues of Apache Kafka for asynchronous, service-to-service messaging.

andersswanson.dev

When Code Is Cheap, Agent Memory Becomes the Moat

Agent memory is becoming one of the most important ways to differentiate an AI application. It's getting much cheaper to turn a good idea into a working app. A small, skilled team with capable models and modern coding tools can get surprisingly far in a couple days. Increasingly, features that took months and a specialized team are faster to prototype, and faster for competitors to copy.

andersswanson.dev

No separate database required: real-time, multi-model data processing

Real-time, multi-model data processing (and many microservices workflows) often involves transactions across several systems: an event stream, relational history, location data, vector search, and more. When a failure occurs, it can be difficult to reason about: Which work committed? Which message was published? What now needs compensating? In this code example, Oracle AI Database handles relational data, transactions, event streaming, spatial queries, vector search, and AI inference from one platform.

andersswanson.dev

Why I’m done with WebFlux & reactive Java

I remember developing my first reactive application years ago, not too long out of college, using Java 8 and Spring WebFlux. When building the app, it was a seemingly simple decision to use reactive programming: our app dispatched many concurrent HTTP requests to downstream webhooks, so a non-blocking, asynchronous framework would provide the best performance. At the time, reactive programming was also…

andersswanson.dev

The Value Proposition of a Multi-Model Database

A multi-model database is a database that handles multiple data types, like vectors, graphs, documents, relational data, and more - in a single, converged engine. Multi-model databases are most valuable when multiple data types are closely related and must be queried transactionally. When working with real-world, multi-model data, I've found this to often be the case.

andersswanson.dev

Select AI on a local database, with Testcontainers

Select AI uses natural-language prompts to generate, run, inspect, and explain your data. Actions such as showsql, runsql, explainsql, and narrate allow you to review a generated query before execution, understand how it was formed, or turn results into a readable response. Select AI runs in the database, close to your data, making it uniquely suited for natural language to SQL workflows.

andersswanson.dev

Connection Pooling: How to Tune Oracle UCP

Connection pool examples usually start with pool sizing and end with a happy-path query. If you're building a production-grade service with Oracle UCP (Universal Connection Pool), that might miss the point: Your connection pool includes a vast array of features, covering failover, database resident connection pooling, caching, and more. In this article, we'll walk through how to configure several common Oracle UCP scenarios in Spring Boot.

andersswanson.dev

Local model inference with Ollama

Sometimes we don't need the most powerful frontier model. Sometimes we just need a local model that's good enough. Depending on the agent harness, local models are also great for data privacy! In this article, we'll use Ollama to install and use gpt-oss:20b for local inference and agent workloads. Install Ollama Ollama is easy to install on most platforms. Simply visit the…

andersswanson.dev

Building Agent Memory: Why your agents need real memory

When we talk about agent memory, we look to compose memory formation, retrieval, and persistent storage. You've probably seen an agent like Claude Code or Codex re-read the same file it just read. Most agents do not have durable memory by default. They have context for the current run, but not persistent, managed experience across runs.

andersswanson.dev

JPA exception handling: two strategies

When repository calls fail, JPA exceptions may be ungracefully thrown, causing your app to return HTTP 500 codes, and fail to appropriately respond to the error. Your app log most likely has the real error code (like an ORA- code) buried in a stack trace, but no action was taken on that error. However, wrapping every repository call in try/catch…

andersswanson.dev

Get More Out of Your Coding Agents (as an Oracle Developer) - andersswanson.dev

Get more out of your coding agents as an Oracle AI Database developer with skills, MCP, and an understanding of best practices.

buff.ly

Deep Data Security: row and column level RBAC

Implementing client-side row and column level RBAC is a fun challenge. But what if you handle everything server-side? That's exactly the problem Deep Data Security solves for Oracle AI Database. Deep Data Security in Oracle AI Database is designed so authorization resides in the database, tied to an end user, their data roles, and runtime context. Policies automatically apply security filters, with no client-side intervention.

andersswanson.dev

It’s time to upgrade to Spring Boot 4.1 (and Java 25)

Still on Spring Boot 3.x, or even 2.7 in production? Spring Boot 4 is more than a framework upgrade: it’s a chance to reduce dependency weight, improve startup behavior, modernize your Java runtime, and selectively evaluate GraalVM for cold-start-sensitive services. This guide lays out a practical upgrade path across Spring Boot 4, Java 25, and native images, with a focus on measurable performance gains and migration risk.

andersswanson.dev

Get More Out of Your Coding Agents (as an Oracle Developer)

Most bad agent output isn't too mysterious. Usually, the agent is missing the proper shape of your system. Your Oracle AI Database work may be split into live schemas, distributed apps, views, grants, SQL dialect details, generated DDL, docs, and the few sentences in your head about what the app actually does. In this article we'll look at using agents intelligently to develop with Oracle AI Database.

andersswanson.dev

Brand new to Oracle AI Database? Start Here - andersswanson.dev

Many Oracle guides are written for DBAs or long-time developers, and as a newcomer it feels like you’re expected to already know the jargon. I started in the same place, learning the as I went, and…

buff.ly

A tour of LangChain Oracle ingestion and retrieval

The langchain-oracledb package makes it easy to integrate LangChain concepts with your Oracle AI Database instance. In this article, we'll explore the langchain_retrieval sample, using langchain-oracledb for content retrieval. The sample composes various langchain-oracledb classes, including to load, split, save, and retrieve content. This post is a companion article to my prior post, LangGraph persistence with Oracle AI Database…

andersswanson.dev

LangGraph persistence with Oracle AI Database

Key Takeaways Oracle AI Database can be used as a persistence layer for LangGraph. Agent workflows need durable state to pause and resume for real decisions. OracleSaver checkpoints graph execution state so the workflow can resume using the same thread_id. A checkpoint preserves workflow state and resume history; an approved record is application data that can be queried or audited later.

andersswanson.dev

Agent Memory Techniques for Oracle AI Database

Key Takeaways Database agent memory should be modeled as durable application data, not just as embeddings in a vector table. Oracle AI Database can combine vectors, relational columns, JSON metadata, Oracle Text, property graphs, and timestamps in one memory system. Hybrid retrieval is stronger than vector search alone because it can match meaning, exact identifiers, entities, relationships, recency, and importance.

andersswanson.dev

Is Oracle the best database for small to mid size shops?

Key Takeaways Oracle AI Database is worth a serious look for small and midsize teams when one application needs relational data, JSON, search, vectors, spatial queries, REST endpoints, and internal app tooling without operating a stack of separate services. In this case, it may be the best database for your needs. The free path lets you verify it: local containers, FreeSQL, SQLcl, SQL Developer, and Always Free Autonomous AI Database let you prove before paying.

andersswanson.dev

Polyglot, automated ORDS tests with Testcontainers

Did you know you can locally test Oracle REST Data Services (ORDS) with free Oracle AI Database resources? If you rely on mock servers (or don't run automated tests at all!), you're dealing with a thin layer of comfort. Using container-driven ORDS tests can prove your app starts, connects to ORDS, and uses database APIs like the MongoDB wire-compatibility API. All with a local stack your test created.

andersswanson.dev

An Agent Skill that uses Kafka Java APIs for Oracle AI Database

In my own work, I found most coding agents weren't generating high-quality code for Oracle AI Database's Kafka Java API (OKafka). You can get results, but they're not idiomatic, and miss subtleties. This is why I created the okafka-java-code oracle agent skill, based off my hand-written Kafka Java API examples. Agent skills can greatly enhance code generation for Oracle AI Database apps, and this skill encapsulates solutions to the problems I kept hand-coding: how to authenticate with OKafka, how to use transactions, how to create topics, and how to use Oracle-specific serialization.

andersswanson.dev

An app that keeps its data in one place : multi-model database pattern

The code for this article on multi-model database patterns lives in the oracle-database-code-samples repo, under the support-ticket-intelligence sample.  When I add a graph database to an application, I'm not just adding a query feature. I need to understand how the database is set up, how its schema works, how to operate it, how to monitor it, and how to move data across service boundaries.

andersswanson.dev

Mongo Over Relational Data: JSON Relational Duality Views

Can your MongoDB client be used against a document-shaped contract while the system of record stays relational? JSON Relational Duality Views expose document APIs over relational data. Those same document APIs accessed through Oracle REST Data Services' (ORDS) MongoDB-compatible API allow any MongoDB client to work with what is essentially transactional, relational data. In this article, we'll implement exactly that, walking through the…

andersswanson.dev

5 Oracle AI Database Dev Tools I’d Put in a Starter Kit

Developers need the shortest path from claim to proof. Can I start a database locally? Can I connect from my app? Can I run my tests against it? Can I inspect the schema without guessing? Can I use it with scripts, agents, and CI? Can I easily move from a laptop to a managed cloud database? In this article, we'll look at tools that help you shorten the feedback loop for the development process you're trying to prove.

andersswanson.dev

Get Started with Spatial Data in Oracle AI Database Using JDBC - andersswanson.dev

Spatial data is a database feature that sounds niche right up until you need it. When your application must answer questions like “what landmarks fall inside this map window?”, “what is the exact…

buff.ly

The Exploit Window is Shrinking. Converged Databases Shrink the Attack Surface

I often talk about converged, multi-model databases from the perspective of developer productivity. Owning one platform that handles many data types like relational, JSON, graph, vector yields fewer moving parts when building and operating apps. Recently, another reason is more and more compelling: security. AI-assisted vulnerability discovery is moving incredibly fast. Anthropic's Project Glasswing describes frontier models finding and exploiting software vulnerabilities at a level that…

andersswanson.dev

Database Per Service with Oracle AI Database PDBs

Using Oracle AI Database Pluggable Databases (PDBs) to give each Spring Boot service its own clear data boundary. Key Takeaways Learn database-per-service design with separate studentpdb and coursepdb PDBs. Each Spring Boot service owns its own schema, tables, JDBC URL, and connection pool. The runner composes student and course data through HTTP APIs, without shared schemas or cross-PDB joins. The project can be run with Testcontainers or manually with Docker and Maven.

andersswanson.dev

Using Agent Skills to develop with Oracle AI Database

Skills are reusable, task-specific workflows for your agents: each skill is a directory centered on a SKILL.md file, with optional scripts,references, and assets packaged together. Skills bring context efficient capabilities to agents in a reliable, repeatable manner. In this article, we’ll look at what skills are, how to install them, and how the open source oracle-db-skills repo can help your agents do real work with Oracle AI Database.

andersswanson.dev