Thoughts and other thingshttps://blog.eduard3v.com/https://blog.eduard3v.com/posts/en/securing-ai-automation-part-1/https://blog.eduard3v.com/posts/en/securing-ai-automation-part-1/Secure AI automation pipelines against API key leaks and data exfiltration. Covers Vault integration, K8s NetworkPolicy, and incident response for LLM workloads: Part 1Thu, 28 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-4/https://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-4/FAQ covering quantization levels, VRAM differences, Kubernetes deployment, and benchmark cadence for local LLM inference engines. Practical advice from production experience , Part 4Mon, 25 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-3/https://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-3/Decision framework for choosing Ollama, vLLM, or llama.cpp based on your workload. Reproduction instructions, engine selection criteria, and deployment guidance , Part 3Fri, 22 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-2/https://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-2/Benchmark results for Ollama, vLLM, and llama.cpp on Llama 3 8B. Throughput data, latency tables, VRAM consumption, and scaling behavior across concurrency levels , Part 2Tue, 19 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-1/https://blog.eduard3v.com/posts/en/benchmarking-local-llm-inference-part-1/Local LLM benchmark comparing Ollama, vLLM, and llama.cpp on Llama 3 8B. Real tokens/sec, VRAM usage, and batching performance data with test harness code , Part 1Sat, 16 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-n8n-ai-automation-part-3/https://blog.eduard3v.com/posts/en/deploy-n8n-ai-automation-part-3/Production-hardening n8n with performance tuning, health checks, load testing, monitoring, backup strategy, and troubleshooting common errors, Part 3Wed, 13 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-n8n-ai-automation-part-2/https://blog.eduard3v.com/posts/en/deploy-n8n-ai-automation-part-2/Production-ready n8n deployment on Kubernetes with Helm-like manifests, webhook security, credential encryption, and AI Agent node setup for local LLMs , Part 2Sun, 10 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-n8n-ai-automation-part-1/https://blog.eduard3v.com/posts/en/deploy-n8n-ai-automation-part-1/Production-ready n8n deployment with Docker Compose and Kubernetes. Covers queue mode, credential encryption, PostgreSQL backend, and AI Agent node setup , Part 1Thu, 07 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-5/https://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-5/Step-by-step deployment guide for n8n and Temporal self-hosted setups. Docker Compose configs, Python SDK example, and production recommendations: Part 5Mon, 04 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-4/https://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-4/When to choose Temporal over n8n for AI workflows, plus hybrid architectures that use both. Includes migration paths and production patterns: Part 4Fri, 01 May 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-3/https://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-3/Deep dive into error handling patterns for n8n vs Temporal. Compare retry policies, idempotency, saga patterns, and production reliability strategies: Part 3Tue, 28 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-2/https://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-2/Detailed feature comparison of n8n and Temporal across 14 dimensions including AI integration, durability, scalability, and cost for AI workflows: Part 2Sat, 25 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-1/https://blog.eduard3v.com/posts/en/n8n-vs-temporal-part-1/Compare n8n and Temporal for AI workflow orchestration. Visual no-code vs code-first durable execution. Find the right tool for your stack: Part 1Wed, 22 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-5/https://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-5/Assess your Agentic DevOps maturity from Level 1 to Level 5 Full Autonomy. Plus FAQs on MCP servers, production safety, and governance: Part 5Sun, 19 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-4/https://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-4/Follow a 7-step path from audit to autonomous operations for Agentic DevOps. Includes MCP server setup, governance patterns, and key metrics to track: Part 4Thu, 16 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-3/https://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-3/Compare Claude Code, LangGraph, and OpenAI Agents SDK for Agentic DevOps. Plus security risks, hallucination guardrails, and governance patterns: Part 3Mon, 13 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-2/https://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-2/Learn how MCP servers connect AI models to infrastructure APIs through Model Context Protocol. Build a Kubernetes log query agent with Python: Part 2Fri, 10 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-1/https://blog.eduard3v.com/posts/en/what-is-agentic-devops-part-1/Agentic DevOps uses AI agents and MCP servers for autonomous infrastructure ops. Learn how the Model Context Protocol powers modern AI SRE workflows: Part 1Tue, 07 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-4/https://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-4/Production hardening for Ollama on Kubernetes: HPA alternatives, NetworkPolicies, monitoring with Prometheus, troubleshooting GPU/OOM errors, and when to avoid Kubernetes, Part 4Sat, 04 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-3/https://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-3/Production-ready YAML manifests for Ollama on Kubernetes: PVC, ConfigMap, Deployment with GPU scheduling, liveness probes, Service, Ingress, and verification, Part 3Wed, 01 Apr 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-2/https://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-2/Architecture design for deploying Ollama on Kubernetes: namespace isolation, resource quotas, GPU node affinity, and ConfigMap patterns for production inference, Part 2Sun, 29 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-1/https://blog.eduard3v.com/posts/en/deploy-ollama-kubernetes-part-1/Step-by-step guide to deploy Ollama on Kubernetes with GPU scheduling, persistent storage, and production YAML manifests. Covers prerequisites and NVIDIA GPU Operator setup, Part 1Thu, 26 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/event-driven-ai-pipelines-part-3/https://blog.eduard3v.com/posts/en/event-driven-ai-pipelines-part-3/Build event-driven AI pipelines with n8n, Redis, Temporal, and LLMs. Idempotency, backpressure, error handling, and observability on Kubernetes, Part 3Mon, 23 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/event-driven-ai-pipelines-part-2/https://blog.eduard3v.com/posts/en/event-driven-ai-pipelines-part-2/Build event-driven AI pipelines with n8n, Redis, Temporal, and LLMs. Idempotency, backpressure, error handling, and observability on Kubernetes, Part 2Fri, 20 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/event-driven-ai-pipelines-part-1/https://blog.eduard3v.com/posts/en/event-driven-ai-pipelines-part-1/Build event-driven AI pipelines with n8n, Redis, Temporal, and LLMs. Idempotency, backpressure, error handling, and observability on Kubernetes, Part 1Tue, 17 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-temporal-ai-workflows-part-3/https://blog.eduard3v.com/posts/en/deploy-temporal-ai-workflows-part-3/Configure persistence, saga patterns for AI, security hardening, Prometheus monitoring, n8n integration, and CI/CD pipelines for Temporal workers on Kubernetes.Sat, 14 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-temporal-ai-workflows-part-2/https://blog.eduard3v.com/posts/en/deploy-temporal-ai-workflows-part-2/Deploy Temporal Server on Kubernetes with PostgreSQL, Elasticsearch, Python workers. Covers secrets, ConfigMaps, HPA scaling, Prometheus metrics for production.Wed, 11 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-temporal-ai-workflows-part-1/https://blog.eduard3v.com/posts/en/deploy-temporal-ai-workflows-part-1/Learn Temporal's architecture, four core services, and production Kubernetes prerequisites. Includes saga patterns, deployment prerequisites for AI pipelines.Sun, 08 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-4/https://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-4/Harden your LiteLLM deployment for production: high availability, auth, rate limiting, monitoring, TLS, and a troubleshooting guide for common issues.Thu, 05 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-3/https://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-3/Connect AI tools to LiteLLM on Kubernetes: OpenCode, Cursor, Continue.dev. Then dive deep into provider configuration for Kimi, OpenRouter, and NVIDIA NIM.Mon, 02 Mar 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-2/https://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-2/Deploy LiteLLM to Kubernetes with production-ready manifests: proxy_config.yaml, Secrets, Deployment, Service, and Postgres setup for your AI gateway.Fri, 27 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-1/https://blog.eduard3v.com/posts/en/litellm-kubernetes-ai-gateway-part-1/Learn how LiteLLM works as a centralized AI gateway on Kubernetes. Architecture overview, prerequisites, and what you'll build across this 4-part series.Tue, 24 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-vllm-production-part-6/https://blog.eduard3v.com/posts/en/deploy-vllm-production-part-6/Conclusion and frequently asked questions about deploying vLLM on Kubernetes: GPU memory requirements, quantization, scaling, HPA, and monitoring.Sat, 21 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-vllm-production-part-5/https://blog.eduard3v.com/posts/en/deploy-vllm-production-part-5/vLLM vs Ollama comparison, 10-item production readiness checklist, and troubleshooting common vLLM issues on Kubernetes.Wed, 18 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-vllm-production-part-4/https://blog.eduard3v.com/posts/en/deploy-vllm-production-part-4/Performance tuning for vLLM: KV cache management, scheduling strategies, GPU kernel warm-up, and monitoring with PromQL queries for TTFT, TPOT, throughput, and queue depth.Sun, 15 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-vllm-production-part-3/https://blog.eduard3v.com/posts/en/deploy-vllm-production-part-3/Production configuration for vLLM: tensor parallelism setup, AWQ and FP8 quantization, batch size tuning, memory limits, and graceful shutdown handling.Thu, 12 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-vllm-production-part-2/https://blog.eduard3v.com/posts/en/deploy-vllm-production-part-2/Step-by-step guide to deploying vLLM on Kubernetes: namespace, ConfigMap, Secret, StorageClass, PVC, Deployment, Service, Ingress, and HPA with custom metrics.Mon, 09 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/deploy-vllm-production-part-1/https://blog.eduard3v.com/posts/en/deploy-vllm-production-part-1/Introduction to vLLM on Kubernetes: architecture, PagedAttention, continuous batching, tensor parallelism, prerequisites, and version compatibility matrix.Fri, 06 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-4/https://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-4/Privacy and security comparison, MCP support, decision framework, use case matrix, and a complete FAQ for Claude Code, GitHub Copilot, and Continue.dev.Tue, 03 Feb 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-3/https://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-3/Three infra scenarios comparing Claude Code, GitHub Copilot, and Continue.dev : debugging Kubernetes, writing Terraform, and reviewing Helm chart security.Sat, 31 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-2/https://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-2/Context windows, multi-file reasoning, agentic capabilities, tool use, MCP support, and terminal integration for Claude Code, GitHub Copilot, and Continue.Wed, 28 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-1/https://blog.eduard3v.com/posts/en/claude-code-vs-copilot-continue-part-1/Compare Claude Code, GitHub Copilot, and Continue.dev capabilities for infrastructure and SRE workflows. Quick verdict, comparison table, and use case overview.Sun, 25 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-4/https://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-4/Cost comparison, community momentum analysis, and final verdict for Ollama vs vLLM self-hosted inference. Includes FAQ and when to use both tools together.Thu, 22 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-3/https://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-3/A decision framework for choosing between Ollama and vLLM for self-hosted LLM inference workloads, plus a step-by-step migration path from Ollama to vLLM.Mon, 19 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-2/https://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-2/Benchmark data comparing Ollama and vLLM throughput, latency, and VRAM usage for LLM inference. Plus Kubernetes readiness, scaling, and operational differences.Fri, 16 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-1/https://blog.eduard3v.com/posts/en/ollama-vs-vllm-part-1/Compare Ollama and vLLM architecture and design for self-hosted LLM inference. Understand PagedAttention memory management, GGUF format, and when each excels.Tue, 13 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/build-ai-sre-agent-mcp-part-3/https://blog.eduard3v.com/posts/en/build-ai-sre-agent-mcp-part-3/Deploy your MCP server into production using Docker and Kubernetes RBAC, explore multi-tool diagnostic workflows, and get answers to common questions quickly.Sat, 10 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/build-ai-sre-agent-mcp-part-2/https://blog.eduard3v.com/posts/en/build-ai-sre-agent-mcp-part-2/Write the FastMCP server code, implement safety guardrails like read-only mode and namespace allowlisting, and connect an SRE agent to Claude Desktop securely.Wed, 07 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/build-ai-sre-agent-mcp-part-1/https://blog.eduard3v.com/posts/en/build-ai-sre-agent-mcp-part-1/Learn the Model Context Protocol (MCP) and the 4-layer architecture for building safe AI SRE agents that diagnose Kubernetes clusters through natural language.Sun, 04 Jan 2026 00:00:00 GMThttps://blog.eduard3v.com/posts/en/meet-the-engineer/https://blog.eduard3v.com/posts/en/meet-the-engineer/Hi, I'm Eduardo, a Software and DevOps Engineer obsessed with automation, infrastructure, AI, and building things that actually solve problems. CKA certified with enterprise experience at NTT DATA and Evolutio.Thu, 01 Jan 2026 00:00:00 GMT