AI use-case intake & review
Once one team has an agent, everyone wants one, and IT tends to find out afterwards. This agent turns a team's request into a proper record: what job it would do, which sources it needs, what it should be allowed to change, and who approves. IT and security review that record before any co-build starts.
Step by step
Take the request from the team
A team describes the job they want an agent to do, in their own words, in the channel.
Ask the governance questions
The agent asks what data it would read, whether personal or regulated data is involved, and what it would need to change.
Map it to the controls
The request is written up against the same controls every agent has: context, applications, rights, escalation and verification.
Link to agents already built
Existing agents and approved tools that cover part of the job are linked, so two teams do not build the same thing.
Draft the review record
A record is drafted for the IT and security reviewers, with the open questions listed rather than answered on the team's behalf.
IT and security approve it
The reviewers approve, amend or decline the request. A co-build starts only after that.
Exactly what this agent can see, touch and change
The same five controls sit behind every Askollo agent. These are this one's settings — visible before you build it, not buried in an admin screen afterwards.
Context
What reaches the modelIT and platform leads being asked for AI by several teams at once, who want a single front door rather than a scattering of unreviewed pilots.
Escalation
When it asks a personIT and security approve the review record before a co-build starts
Applications & Rights
Which tools it uses, and what it may do in each
NotionRead only
SlackRead only
ConfluenceReadWrites newCreates new records or documents. Never edits, moves or deletes anything that was already there.Verification
How you know it’s rightEvery claim links to the document it came from. A statement the agent cannot cite does not make it into the output — which is what makes the result reviewable in minutes rather than re-read end to end.
Who it’s for
IT and platform leads being asked for AI by several teams at once, who want a single front door rather than a scattering of unreviewed pilots.
What you’ll need
- A place to keep the register of agents
- Your data classification rules
- Named IT and security reviewers
- An agreed review record format
What you get
- A structured request record
- A mapping to rights and gates
- A list of open questions for review
What it doesn’t do
It gathers and records. It does not approve a use case, decide whether data may be used, or build the agent, and it cannot tell you whether a request is worth doing.
We build the first one with you
Not a template you configure alone. We sit with your team, build it on real data, and hand over the controls.
Scope
One session with the people who actually do the work. We agree what the agent reads, what it may write, and who approves.
Co-build
Built on your own data, not a sandbox. You watch it being made, so you know why it behaves the way it does.
Handover
You own the controls. Change the context, tighten the rights, move the approval gate — without coming back to us.
Parts of the work currently spread across the categories below. It does not replace any of those products outright.
Frequently asked
Where does this sit against our existing change process?
In front of it. The record is drafted so your reviewers receive a complete request in the format they already use, and the approval stays in your process rather than moving into ours.
Does it decide which model or cloud a use case runs on?
No. It records what the team is asking for and what constraints apply, such as data that must stay in a particular region. Your reviewers decide, and the decision is written into the record.
What if a team goes ahead anyway?
The agent cannot prevent that, and no software can. What it offers is a route that is quicker than going around it, plus a register showing what has been asked for and what was decided.
Is the register visible to the whole company?
That is your call, and it is set by rights on the source. Many teams make the list of agents visible so people reuse what exists, while keeping the review detail to IT and security.