Managed by Alian Software
AI

AI Adoption Roadmap: A 30/60/90-Day Plan Your Team Can Follow

A practical AI adoption roadmap for teams: what to switch on in the first 30, 60 and 90 days with AlianHub agents, approvals and budgets, and what to measure.

AlianHub Team · 5 min read

An AI adoption roadmap is a staged plan for bringing AI into a team's daily work: start with reading and asking, move to proposals, and only then let agents act on their own. This 30/60/90-day plan uses AlianHub's own AI features, Ask, agents, autonomy levels, the AI Inbox and spend caps, so each stage has a clear switch and a clear way to check it worked.

Key takeaways

  • Days 1-30: let people ask. Days 31-60: let agents propose. Days 61-90: let trusted agents act.
  • Every stage has an off switch. You can always pause all agents in one click.
  • Measure with your own data: approval rate, error rate and cost per agent.
  • One person should own the AI Inbox from day one.
  • Roll out to one team first. Copy what works.

Why most AI rollouts stall

Teams usually fail at AI in one of two ways. Either they switch everything on at once and someone gets burned, or they run a pilot that nobody uses and it quietly dies.

A roadmap fixes both. It gives people small wins early, and it holds back the risky parts until you have evidence.

Trust in AI is built in weeks, not in a kickoff meeting.

The plan at a glance

PhaseGoalAlianHub featuresAutonomy
Days 1-30Get people askingAsk, AI memory, AI project generatorNone, people only
Days 31-60Agents proposeAgents at L0/L1, AI Inbox, audit logL0 and L1
Days 61-90Trusted agents actL2/L3 for low-risk jobs, schedules, team packsUp to L3 for proven agents

Days 1-30: ask, don't act

The first month is about habits, not automation. Nobody's work changes without them.

What to switch on

  • Ask. People ask plain-word questions about projects they can open. Starters like "Find my urgent work" and "Project status update" help them get going.
  • AI memory. Each person tells AI how they like to work, such as short answers or UK spelling. It is private, admins included.
  • AI project generator. For new projects, write a brief and review the full plan before anything is built.

How to do this in AlianHub

  1. As the owner, set the AI provider, key and model under Settings › Instance › Settings, in the AI group.
  2. Set a workspace daily AI budget.
  3. Ask each person to open AI in the left menu and try three questions in Ask this week.
  4. Show them Your AI memory, with "Use my memory" and "Forget everything".
  5. Optionally, let people connect their own AI under AI, then Connect your AI.

Success looks like: people use Ask without being reminded.

Days 31-60: agents propose

Now introduce agents, but keep every change behind a person.

What to switch on

  • One or two agents with simple skills, such as digest.ceo for a weekly digest or brief.parse for breaking briefs into subtasks.
  • The AI Inbox, owned by one named person per team.
  • The audit log, reviewed weekly.

How to do this in AlianHub

  1. Open AI, then the Agents tab. Pick one agent per team.
  2. Set it to L0 or L1, and give it a monthly spend cap and daily action limit.
  3. Assign it a few tasks the way you would assign a person.
  4. Each day, the inbox owner opens AI Inbox and chooses Approve or Decline, adding a reason on declines.
  5. Each Friday, open Audit log, filter to "Agents only", and note what was approved, declined or refused.

Tip: decline reasons are your training data and your documentation. A clear reason helps the agent and tells the next reviewer why the call was made.

Success looks like: most proposals for a given agent are approved with little editing.

Days 61-90: trusted agents act

Only now do agents act without asking, and only for low-risk work.

What to switch on

  • L2 for agents whose low-risk proposals you approve almost every time. Comments, nudges, status changes within a list and reports will run on their own, with Undo.
  • L3 for scheduled jobs, like a Monday digest.
  • Team packs for more teams. Agents start paused and the whole pack can be undone.

How to do this in AlianHub

  1. Check each agent's approval and error rate in the AI health view on the AI controls screens.
  2. Raise proven agents to L2, or L3 for scheduled work. The agent's revision history records the change.
  3. Keep sensitive projects on "Propose everything" so they stay proposal-only.
  4. Switch on a team pack for the next team and repeat days 31-60 with its agents.
  5. Make sure an owner knows how to Pause all agents in a project, under "Agents working at the same time".

Success looks like: routine updates happen on their own, and the AI Inbox only holds decisions worth a person's time.

A business scenario

A professional services firm rolls this out in its delivery team first. In month one, consultants use Ask to prepare for client calls. In month two, a Reporter agent proposes weekly digests and the delivery lead approves them. In month three, the digest runs on its own and the firm switches on packs for finance and sales.

The result: fewer missed deadlines because stuck work shows up every Monday, faster client reporting, and a rollout the partners trust because every step is in the audit log. The firm grows its AI use team by team, with no big-bang risk.

What to measure

Use your own numbers, not industry averages:

  • Approval rate per agent. Rising means it is learning your standards.
  • Error rate per agent and per model.
  • Cost per agent, per model and per feature, against your spend caps.
  • Refusals in the audit log. These show agents trying actions they cannot take.
  • People's habits. Are they still asking after the first week?

All of these are on the AI controls page. For the theory behind each level, read AI agent autonomy levels.

FAQ

What is an AI adoption roadmap?

It is a staged plan for introducing AI to a team. This one moves from asking (days 1-30) to proposing (31-60) to acting on low-risk work (61-90).

Can we go faster than 90 days?

Yes, if your evidence supports it. Move an agent up a level when its approval rate is high, not just because the calendar says so.

What if something goes wrong mid-rollout?

Changes made on their own can be undone from the audit log, and an owner can pause all agents in a project or across the workspace in one click.

Do all teams need to start at once?

No. Start with one team, then switch on team packs for others once the first team's agents have earned trust.

Start your first 30 days on the live demo or a free self-hosted install, and share this roadmap with the people planning your team's AI rollout. Questions? Get in touch.

  • #AI adoption roadmap
  • #AI rollout plan
  • #AI agents
Try AlianHub on your own serverFree and open source under AGPL-3.0. No per-seat pricing.
Explore the Live Demo
Wait! Before you go…

Try AlianHub freefor 7 days

Get your own private AlianHub, hosted and managed by us. Start the free trial on the Starter plan and bring your whole team.

  • Every feature included, nothing locked
  • Unlimited users and guests
  • Hosted, secured and updated for you
Start my free trial

7-day free trial on the Starter plan

View Brochure