Practical AI agents, local LLM experiments, automation and engineering notes from Fazal Shah.https://fazalshah.com/en-ushttps://fazalshah.com/blog/agent-human-approval-design/https://fazalshah.com/blog/agent-human-approval-design/Design approval requests with clear payloads, reviewers, expiry and safe recovery instead of a vague approve button.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/agent-retries-idempotency/https://fazalshah.com/blog/agent-retries-idempotency/Design retry budgets, logical action identifiers and reconciliation for agents calling external systems.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/ai-agents-seo-automation/https://fazalshah.com/blog/ai-agents-seo-automation/Design an SEO automation workflow that collects crawl evidence, drafts fixes, validates output and measures changes without invented ranking gains.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/ai-automation-consulting-pricing/https://fazalshah.com/blog/ai-automation-consulting-pricing/Understand discovery, implementation and operating costs with transparent worked examples and a practical automation scoping checklist.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/ai-workflow-automation-tools/https://fazalshah.com/blog/ai-workflow-automation-tools/Compare n8n, Flowise and custom code using workflow ownership, approval boundaries, deployment needs and a practical acceptance test.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/csv-to-json-automation/https://fazalshah.com/blog/csv-to-json-automation/Handle quoted fields, leading zeros, duplicate headers and explicit types when converting CSV datasets into JSON workflows.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/evaluate-ai-agent-reliability/https://fazalshah.com/blog/evaluate-ai-agent-reliability/Build outcome-based agent tests, separate safety from task success, and understand pass-rate uncertainty with a worked dataset.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/llm-benchmark-latency-throughput/https://fazalshah.com/blog/llm-benchmark-latency-throughput/Our M2 Max 64GB oMLX capture shows 81.0 tok/s solo and 213.1 tok/s aggregate at 4× batching, while E2E latency rises from 2.662 to 8.526 seconds.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/llm-context-memory-bandwidth/https://fazalshah.com/blog/llm-context-memory-bandwidth/Understand weight capacity, KV-cache growth and Apple silicon bandwidth without turning hardware specifications into token-speed promises.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/local-llm-on-mac/https://fazalshah.com/blog/local-llm-on-mac/Plan local AI on Apple silicon, choose a starting interface, budget memory and verify that your intended workload remains local.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/local-vs-cloud-ai-costs/https://fazalshah.com/blog/local-vs-cloud-ai-costs/Compare capital cost, API usage, operator time and task quality with transparent assumptions rather than a universal cheaper option.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/mlx-vs-ollama-mac/https://fazalshah.com/blog/mlx-vs-ollama-mac/A documented MLX LM and Ollama comparison covering model artifacts, integration needs, memory planning and a fair benchmark protocol.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/monitor-ai-agents-production/https://fazalshah.com/blog/monitor-ai-agents-production/Design agent traces, meaningful service metrics, failure alerts and a recovery runbook without confusing model output with completed work.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/n8n-production-runbook/https://fazalshah.com/blog/n8n-production-runbook/Prepare ownership, credentials, execution recovery and acceptance tests for an operational n8n deployment.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/production-agent-boundaries/https://fazalshah.com/blog/production-agent-boundaries/A design note on permissions, approval gates and failure recovery for tool-using agents.Mon, 12 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/rag-evaluation-dataset/https://fazalshah.com/blog/rag-evaluation-dataset/Create a document-grounded evaluation dataset that separates retrieval quality, answer support and safe refusal.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/reproducible-local-llm-tests/https://fazalshah.com/blog/reproducible-local-llm-tests/A measurement protocol for comparing local inference without inventing benchmark results.Mon, 12 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/secure-ai-agents-production/https://fazalshah.com/blog/secure-ai-agents-production/A practical agent security design with tool permissions, exact-action approvals, untrusted data handling and failure-path tests.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/single-page-seo-audit-method/https://fazalshah.com/blog/single-page-seo-audit-method/Understand what a public HTML audit can establish and how to turn its findings into a prioritized implementation plan.Wed, 14 Oct 2026 00:00:00 GMThttps://fazalshah.com/blog/structured-json-model-output/https://fazalshah.com/blog/structured-json-model-output/Build a model-output contract with schema validation, semantic checks and safe failure handling before integrating writes.Wed, 14 Oct 2026 00:00:00 GMT