What's stopping your MSP from scaling with AI?
Most AI pilots stall for one reason: the shop underneath isn't ready for agents to run the work. This assessment measures the thing that actually decides it — whether your data, processes, automation, skills, tooling, and governance are standardized enough for agents to take over the repeatable work, safely. You get your readiness level, the one thing holding you back, and what to fix first.
What it measures — the 6 dimensions
You rate each on how it is today (a mirror, not a grade). For every one, the app already holds the “what good looks like” standard you can browse and adopt.
Illustrative — each assessment scores you across these 6 dimensions on a 1–5 scale; the dent marks your lowest, the constraint to fix first.
-
Data & Documentation Readiness
Can an AI agent ground on your data? Is your environment and service knowledge documented and structured — or locked in people's heads?
What good looks like · A structured, documented backbone — records, runbooks, KB, and CMDB — that an agent can retrieve and reason over. -
Process Standardization
Are your workflows defined and repeatable, so an agent can execute them the same way every time — not improvised per technician?
What good looks like · Defined workflow recipes + statuses + closure codes that an agent can follow deterministically. -
Automation Leverage
Is recurring work automated at the source, or re-solved by hand each time — so growth means hiring rather than capability?
What good looks like · A maintained event→action automation library removes recurring work at the source, and agents draft the rest. -
AI Skills & Team
Does your team have the AI/automation skills — prompt engineering, LLM ops, agent integration, RAG, MCP, AI safety — to build and run agentic workflows?
What good looks like · A team with the modern AI-ops skill set, mapped to roles, that can build, evaluate, and operate agents. -
AI Tooling & Integration
Is AI/automation tooling in your stack, and is your PSA + data integration-ready (APIs, MCP) so agents can actually act, not just chat?
What good looks like · AI/automation tooling in the stack and integration-ready systems (APIs / MCP) agents can read and write through. -
AI Governance & Safety
Do you govern AI use responsibly — safety guardrails, human-in-the-loop on destructive actions, data privacy in AI pipelines, and an AI-use policy?
What good looks like · An AI-use policy, output validation + PII handling, and human-in-the-loop gates on destructive actions — adoption you can defend to a client or auditor.
The 5 readiness levels
Where you land is your headline level; your lowest-rated dimension is the constraint holding you back — you're only as repeatable as your weakest critical area.
-
Level 1 — Ad-hoc
AI is unused or one-off. Work is done by hand ticket-by-ticket; nothing about the operating model is structured for an agent to read or act on.
-
Level 2 — Exploring
Isolated experiments — a chatbot here, a few RMM scripts there. No standardization, and the data an agent would need to ground on is scattered and unstructured.
-
Level 3 — Operationalizing
AI and automation are embedded in some workflows. Core processes and environment data are documented enough to ground on, and the team is deliberately building AI skills.
-
Level 4 — Scaling
Agents handle recurring work across the operating model — triage, dispatch, resolution drafting, documentation — on a standardized, governed, measured footing.
-
Level 5 — Autonomous
AI agents run the repeatable work end-to-end with human oversight by exception. Throughput scales with capability, not headcount, and the system improves continuously.
Why it matters
Agents don't fail because the model is weak; they fail because the operation underneath them is unstructured — tribal knowledge an agent can't read, processes that change per technician, no guardrails on destructive actions. The fastest path up is rarely “more AI”; it's structuring your data, standardizing your processes, and governing the rollout so agents can act. This shows you exactly which of those is your bottleneck.
What you get
- Your headline maturity level across all 6 dimensions.
- The single constraint holding you back — and the targeted next-steps to lift it, cross-linked to the reference.
- An evidence check: where your MSP Profile contradicts a high self-rating, the assessment challenges it (it never overwrites your answer).
- A branded PDF report to download and share with your team.
The rest of the suite
Four honest, size-aware self-assessments — diagnostics, not badges. Run the others to see the full picture.
See where your MSP stands — in minutes.
No install, free to explore. Pick an assessment and get a tailored read on your gaps and your maturity.
Open the reference →