We build agentic software, use AI agents to develop software, and lead the teams doing both. Hands-on AI engineering, development workflows, and technical leadership.https://www.mnlt.ai/https://www.mnlt.ai/blog/complexity-is-easy-reliability-takes-design/https://www.mnlt.ai/blog/complexity-is-easy-reliability-takes-design/Adding agents, tools, and autonomy doesn't make AI systems reliable. Here's what the evidence shows, and the ten design controls that make agents dependable.Tue, 06 Oct 2026 00:00:00 GMTai-agentsreliabilityagent-architectureai-engineeringhttps://www.mnlt.ai/blog/graph-engineering-for-ai-agents/https://www.mnlt.ai/blog/graph-engineering-for-ai-agents/How to design multi-agent workflows as graphs: remove waits that move no data, give nodes schemas, verify against anchors, and know when one agent wins.Tue, 06 Oct 2026 00:00:00 GMTai-agentsmulti-agent-systemsagent-architectureclaude-codehttps://www.mnlt.ai/blog/how-to-design-evals-for-ai-agents/https://www.mnlt.ai/blog/how-to-design-evals-for-ai-agents/An eight-step method for designing AI agent evals: build tasks from real failures, pick graders, report pass^k, calibrate LLM judges, and gate releases in CI.Tue, 06 Oct 2026 00:00:00 GMTai-agentsevalsllm-as-judgeai-engineeringhttps://www.mnlt.ai/blog/right-size-your-ai-spend/https://www.mnlt.ai/blog/right-size-your-ai-spend/Most AI work doesn't need a frontier model. Find each workload's threshold with evals, route to the cheapest model that passes, and know when not to.Tue, 06 Oct 2026 00:00:00 GMTllm-costmodel-routingevalsai-strategyhttps://www.mnlt.ai/blog/github-august-outage-retry-storms/https://www.mnlt.ai/blog/github-august-outage-retry-storms/GitHub was degraded for 7h47m on August 17, 2026. A scaling gap started it; retries kept it going. What it means for your CI, your SLA, and your AI clients.Thu, 03 Sep 2026 00:00:00 GMTreliabilityincident-analysisgithubretriesai-engineering