Blog

Engineering notes on AI agents, automation, and the infrastructure behind them.

AI Agents: Build vs Buy (2026 Decision Framework)

AI Agents: Build vs Buy (2026 Decision Framework)

April 11, 2026 · 7 min read
Should you build a custom AI agent or buy off-the-shelf? Cost analysis, hybrid approach, and when to hire a freelancer vs building in-house. Download the free AI Automation Checklist.
Production AI Agent Architecture: Patterns That Actually Ship

Production AI Agent Architecture: Patterns That Actually Ship

April 10, 2026 · 18 min read
Production AI agent architecture patterns: router-planner-executor, ReAct, Reflexion, tool design, state, and testing. The playbook for agents that survive real traffic.
Claude API Pricing Tiers and Cost Optimization Playbook (2026)

Claude API Pricing Tiers and Cost Optimization Playbook (2026)

April 9, 2026 · 10 min read
Claude API pricing tiers, rate limits, and every lever for cutting Anthropic API cost without losing quality. Ten optimizations ordered by ROI. Download the free AI Automation Checklist.
Hetzner vs AWS for AI Workloads: The Honest Breakdown (2026)

Hetzner vs AWS for AI Workloads: The Honest Breakdown (2026)

April 6, 2026 · 15 min read
Hetzner vs AWS for AI workloads in 2026: where each wins, price-per-spec comparison, GPU reality, egress traps, and a realistic decision framework. Download the free AI Automation Checklist.
GPU Cloud Comparison for AI Inference: 2026 Reality Check

GPU Cloud Comparison for AI Inference: 2026 Reality Check

April 4, 2026 · 13 min read
GPU cloud comparison for AI inference in 2026. Lambda, RunPod, CoreWeave, Hetzner, AWS, GCP head-to-head on price, features, and workload fit. Download the free AI Automation Checklist.
Migrate OpenAI to Claude: API Migration Guide for 2026

Migrate OpenAI to Claude: API Migration Guide for 2026

April 4, 2026 · 13 min read
How to migrate OpenAI to Claude: API deltas, feature map, model selection, side-by-side code, cost math, and a 2-week migration plan that actually ships. Download the free AI Automation Checklist.
Qdrant vs Pinecone vs Weaviate: Production Vector DB Comparison 2026

Qdrant vs Pinecone vs Weaviate: Production Vector DB Comparison 2026

April 3, 2026 · 16 min read
Qdrant vs Pinecone vs Weaviate compared on filtering, hybrid search, quantization, and cost. Opinionated verdict for self-hosted and managed RAG. Download the free AI Automation Checklist.
ChatGPT vs Claude Vergleich: Welche KI für DACH-Teams 2026

ChatGPT vs Claude Vergleich: Welche KI für DACH-Teams 2026

April 3, 2026 · 17 min read
ChatGPT vs Claude Vergleich 2026: Features, Kosten, DSGVO und Entwickler-Erfahrung. Klare Empfehlung, welches Modell für welchen Workload passt. Jetzt kostenlose KI-Automatisierungs-Checkliste downloaden.
Teedian: AI Content Operations Engine for Make.com, n8n, Notion

Teedian: AI Content Operations Engine for Make.com, n8n, Notion

April 2, 2026 · 9 min read
Teedian is an AI content operations engine that orchestrates briefings, drafts, approvals, and cross-posting through Make.com, n8n, and Claude. Download the free AI Automation Checklist.
MCP vs Custom API Integration: When Each One Actually Wins

MCP vs Custom API Integration: When Each One Actually Wins

April 2, 2026 · 11 min read
MCP vs custom API integration, decided by real production tradeoffs. When to use an MCP server, when to stay with tools in your Claude code, and how to choose. Download the free AI Automation Checklist.
AI GitHub Issue Triage with FixClaw: Human in the Loop

AI GitHub Issue Triage with FixClaw: Human in the Loop

March 28, 2026 · 9 min read
FixClaw is an AI GitHub issue triage tool. Detects duplicates, classifies severity, drafts labels and comments. Maintainer approves every action. Download the free AI Automation Checklist.
Claude API Tool Use: input_schema, stop_reason, Call Loop

Claude API Tool Use: input_schema, stop_reason, Call Loop

March 26, 2026 · 17 min read
Claude API tool use reference. input_schema (JSON Schema 2020-12), stop_reason flow, parallel calls, streaming, MCP. TypeScript from 10 agents.