Software Architect · AI Agent Security · Distributed Systems
I work on AI agent architecture and security, with a background in distributed systems, JVM engineering, and real-time data platforms. On the blog I write about the trade-offs and failure modes that show up when these systems reach production
- AI agents and security: Agent runtimes, memory and context engineering, authorization, tool security, and evaluation
- Distributed systems and API:s Cloud-native services, event-driven architecture, resilience, and practical API design
- Streaming and data platforms: Kafka, CDC, Spark, Flink, and real-time analytics
- JVM internals and performance: Java, Scala, concurrency, profiling, and runtime optimization
| Project | Description |
|---|---|
| traffic-replay | Replay HTTP traffic from Kafka into test environments for load testing and system validation |
| spark-sketches | Approximate distinct counting in Apache Spark using probabilistic DataSketches |
| otel-playground | Hands-on experiments with OpenTelemetry and manual instrumentation |
| security-in-action-natter-api | Examples of API security, access control, and cryptography |
| spark-bin-packing-partitioner | Bin-packing approaches to reducing data skew in Spark |
Explore all repositories at github.com/gelerion
- Beyond Conversation History: Building Reliable Memory for AI Agents: Designing a runtime memory layer that carries useful knowledge across sessions
- OpenTelemetry Foundations: What Is a Signal?: Understanding traces, metrics, logs, and how they fit together
- Pragmatic API Design: Mastering Collections: Designing collection APIs that stay usable as systems grow
More articles and engineering notes at gelerion.github.io

