Recent content on My Battles With Technology
How I used Claude Code's CLI to audit my DigitalOcean infrastructure, discovered orphaned resources costing me money, and built a plan to slash my cloud bill by 63%. Part 1 of a 6-part series.
A deep dive into mitmproxy for security testing. Learn how to build a scriptable regression harness to catch header drift and token leakage in mobile and API traffic.
Not productivity theater — real scripts I run daily, with before/after time savings, what failed, and exactly why each one earned its place in my toolkit.
Kubernetes Secrets aren't actually secret by default. Here's a practical guide to real secrets management: from etcd encryption to External Secrets Operator and Sealed Secrets.
AI is changing technical work. Not replacing it. Here's an honest look at what's shifting, what stays human, and how to position yourself for the next five years.
Containers aren't magic security boundaries. Here's a practical breakdown of the controls that actually reduce risk: non-root users, read-only filesystems, capability drops, image scanning, and proper secrets handling.
AI changes debugging from pattern-matching against old answers to a genuine diagnostic conversation about your specific problem. Here's a framework that actually works.
Documentation is everyone's least favorite task. AI changes the economics entirely, but introduces new challenges around trust, maintenance, and the human role in technical writing.
The AI tool landscape is overwhelming. Here's a practical framework for evaluating tools, building a focused toolkit, and avoiding the trap of chasing every new release.
Code review is where AI collaboration gets interesting. Learn a practical framework for using AI as a first-pass reviewer while keeping human judgment where it matters most.
Practical solutions for writing scripts that work across Windows, macOS, and Linux. Learn the common pitfalls, workarounds, and patterns that save hours of debugging.
Learn to measure what matters in human-AI collaboration. Move beyond vanity metrics to honest assessment of speed, quality, and growth. Includes the Centaur Scorecard framework.
Learn to calibrate your trust in AI suggestions. Develop verification strategies that catch errors without killing productivity, and build your personal protocol for human-AI collaboration.
Apply the Centaur framework to security tooling. Learn how to build AI-augmented security tools while keeping humans in control of high-stakes decisions.
Move beyond basic autocomplete to true human-AI collaboration. Learn the four collaboration modes that separate Centaur developers from those just along for the ride.
Learn to build a production-grade network port scanner in Python from scratch, covering TCP connect scanning, SYN scanning, service detection, multi-threading, and ethical considerations.
Download the free companion resources for The Centaur's Edge.
Practical guide to security automation scripting with Python and Bash, covering log analysis, vulnerability scanning, incident response automation, and API integration with working examples.
Deep dive into advanced Linux networking including network namespaces, policy-based routing, traffic shaping, VXLANs, and eBPF for network observability and control.
Explore the ethical challenges of using AI in cybersecurity, including algorithmic bias, privacy concerns, adversarial use, and frameworks for responsible AI security implementation.
Learn how to leverage AI and machine learning to analyze security logs, detect anomalies, identify threats, and automate log analysis workflows using Python and open-source tools.
A technical deep dive into using AI and machine learning for automated incident response, covering threat detection, containment workflows, and remediation automation with practical examples.
Learn how to leverage AI and machine learning to analyze log files for security threats. Explore practical techniques using Python, scikit-learn, and LLMs for anomaly detection and threat hunting.
A hands-on account of migrating a technical blog from WordPress to Hugo, including theme selection, content migration challenges, and CI/CD automation lessons learned.
Examine open-source AI model implications: Apache 2.0 vs restrictive licenses, deployment patterns, security risks, and business impact using Llama 2, Mistral, and other major releases as case studies.
Balance AI productivity with critical thinking using the Centaur Mindset framework. Strategies for maintaining cognitive skills while leveraging AI assistance effectively.
Analyze AI-generated code security: vulnerability patterns, data poisoning attacks, bias detection, and secure implementation strategies for Copilot/ChatGPT-generated code.
Execute Gemini 2.5 Pro locally via CLI: 1M token context window, 60 RPM limit, Apache 2.0 license. Covers terminal integration and natural language prompting.
Production WebAuthn/FIDO2 passkey implementation with complete registration/authentication flows, server-side verification, cross-platform syncing, and security analysis. Includes Node.js, Python backends with working code.
Build production-grade AI-powered vulnerability scanning systems using OpenAI GPT-4 API. Includes complete code examples, CI/CD integration, batch scanning workflows, and security best practices.
Implement Anthropic MCP with Azure Blob Storage: client-server architecture, RBAC policies, Redis caching, SSE streaming, and Prometheus monitoring for AI-data integration.
Reduce AI chatbot energy costs with codecarbon metrics, GPT-4/Gemini benchmarking, model quantization, caching strategies, and carbon-aware deployment optimization.
Scan AI-generated code with Endor Labs: integrate SAST/DAST for Copilot/Cursor output, GitHub Actions workflows, custom regex rules, and pre-commit hooks for DevSecOps.
Build production-grade resilient distributed systems with circuit breakers, service mesh, chaos engineering, distributed tracing, and AI-powered failure prediction. Complete implementation guide with Resilience4j, Istio, and OpenTelemetry.
Fabricate PCBs with fiber lasers: 20W laser etching, copper trace cutting, solder mask application, and SMD component assembly. Includes safety protocols and KiCad integration.
Boost productivity with GitHub Copilot for VS Code: AI-powered code completion, natural language prompts, prompt engineering, and security best practices for developers.
Deploy GenAI pipelines on Azure with Azure OpenAI Service, prompt engineering, model monitoring, AKS deployment, and CI/CD integration. Complete working examples with cost optimization.
Transform e-commerce with AI-powered product recommendations, automated content generation, personalized search, and conversion optimization using collaborative filtering and LLMs.
Enhance vulnerability discovery with AI-powered fuzzing: reinforcement learning input generation, coverage-guided testing, and automated exploit development techniques.
Implement anomaly detection with scikit-learn: Isolation Forest algorithm, contamination tuning, threshold optimization, and production deployment patterns with example code.
Explore decentralized AI architectures: federated learning, blockchain model registries, distributed inference, and privacy-preserving computation using real-world frameworks like Ocean Protocol and Flower.
Advanced anomaly detection with deep learning using PyTorch autoencoders, variational autoencoders (VAE), and LSTM networks. Includes production deployment patterns and decision frameworks for choosing between classical ML and neural approaches.
Implement Explainable AI with SHAP, LIME, and saliency maps. Covers model interpretability, feature attribution, and PyTorch/scikit-learn integration techniques.
Build production-grade AI chatbots with LLMs, RAG, and vector databases: GPT-4, LangChain, Pinecone integration, streaming responses, conversation memory, and FastAPI deployment.
Implement WebSockets for real-time communication using WS library, Node.js, and Socket.io. Covers bidirectional messaging, connection handling, and event broadcasting.
Implement Gemini text embeddings for semantic search, RAG applications, and vector similarity matching. Complete examples with Google Generative AI API, vector databases (Pinecone, ChromaDB), and production deployment patterns.
Develop custom applications for Flipper Zero using the Flipper Application Framework. Learn GUI development, GPIO control, event handling, and establish a foundation for building security research tools.
Production security workflows for AI-generated code using Semgrep, Bandit, CodeQL, and secure prompting strategies. Includes complete CI/CD integration, automated scanning, and best practices for GitHub Copilot/ChatGPT/Claude.
Production reinforcement learning with Stable-Baselines3, RLlib, and Gymnasium. Complete DQN, PPO, A2C implementations with custom environments, distributed training, hyperparameter tuning, and deployment patterns.
Detect and remove stalkerware: identify suspicious app permissions, network traffic analysis, battery drain patterns. Covers Android/iOS detection tools and forensic techniques.