I added a AI Scala learning companion to my IntelliJ workbench
One button turns my open Scala file into a focused lesson. How I built a native IntelliJ panel with Codex, connected DeepSeek, and used an unfinished class as the first real example.
Field Notes
Notes on event-driven Java, Spring Boot, Kafka, and secure systems for regulated banking.
One button turns my open Scala file into a focused lesson. How I built a native IntelliJ panel with Codex, connected DeepSeek, and used an unfinished class as the first real example.
What happens when an AI model returns typed decisions and probabilities instead of generating an answer? A look at Jev, calibration, and a possible fast path inside the agent runtime.
A real IntelliJ development instance, a Scala-focused plugin, and my own graphite theme and hoodie mascot. The architecture, desktop screenshots, and tests behind the first working milestone.
Scala and the JVM · Part 3
From Java Optional to Scala Option: represent presence with Some, absence with None, and handle both outcomes with a match expression using our runnable policy lookup.
Scala and the JVM · Part 2
From a Java-style policy class to Scala 2 case classes: constructor properties, value equality versus identity, and copying data without changing the original.
Scala and the JVM · Part 1
Starting Scala 2 from Java and Spring: val and var, expressions, a first method, and a small sbt project you can run. The first step in the Scala/Play series.
I chose a flat ZB monogram, rebuilt it as SVG, and carried it across my portfolio, blog and FOCUS headers. Here are the design choices, the shared-asset surprises, and the checks behind the release.
Cross-entropy looked difficult to me. Let’s unpack −ln(p) from one correct token, understand every symbol, and verify the maths in PyTorch.
Security and privacy, in practice
My Keycloak broker showed a username/password form beside LinkedIn even though LinkedIn was the only supported recruiter identity. Here is how I replaced that ambiguous page with a fail-closed browser flow.
LLMs, the whole thing · Part 10
Chapter 4 is the moment attention becomes a complete GPT. I build the model one inspectable component at a time, count its parameters, and trace the gloriously untrained first token.