Skip to main content
AIX Automation Lab
Book a 30-min discovery call
Menu

Insights

Field notes on building AI that ships.

No hot takes on the latest model release. These are the durable pieces — how we scope, what we build, and the tradeoffs we name on the discovery call. Written for SMB and public-sector operators by the team that runs an MSP.

  1. The first 90 days of AI in an understaffed city IT department

    A 90-day AI plan for city IT: what to ship first, what it actually costs, what to tell the council, and what the CPRA means for your prompts.

    13 min read

  2. What an AI agent actually is — and isn't — for an SMB

    A plain-English definition of AI agents for small business, the three-signal test for when to build one, and when an automation is the smarter buy.

    6 min read

  3. n8n vs Zapier vs custom: when each automation wins

    An operator's n8n vs Zapier vs custom decision guide for SMB and public-sector ops — cost, data residency, and who owns it after the consultant leaves.

    9 min read

  4. RAG for organizations with under 1,000 documents

    An honest guide to RAG for small business: when retrieval-augmented generation is worth building under 1,000 documents, and when good search wins.

    11 min read

  5. We built AVA. Here's the architecture.

    How we architected AVA, the AI assistant our MSP runs on Autotask — the retrieval layer, the bounded action tools, the eval harness, and the cost controls.

    12 min read

  6. Microsoft Bookings + AI: what's actually possible

    An operator's guide to adding AI and automation around Microsoft Bookings — intake triage, reminders, routing, Power Automate, and what's still overhyped.

    11 min read

  7. Autotask + AI: the integration patterns we ship

    Autotask AI integration for MSPs: ticket triage, RAG over ticket history, drafted notes, time-entry help, and the REST API realities behind it all.

    11 min read