Consulting · Readiness assessment
AI agent readiness assessment: find out where agents can safely take work off your team.
You think AI agents could take work off your team, but cannot yet say where, how safely, or in what order.
- A ranked list of the tasks worth handing to an agent
- A review of whether your systems and data are ready, and what must change first
- A pilot, defer or stop decision for each task, in a report you keep
Tell us what you want to automate. We reply by email.
An AI agent readiness assessment shows whether an organisation can run agents safely in production, and where to start. Aelsha Innovations agrees the assessment’s scope and schedule with you at kick-off. It produces a scored list of use cases, a review of data and systems, a 90-day roadmap and a pilot, defer or stop recommendation. You keep the written output.
Technical detail: standards, methods and what we check ↓Terms used on this page
- AI agent
- software that reads a request, decides what to do and acts in your systems.
- Pilot / production
- a pilot shows it can do the task once; production is when it runs on real work every day.
- Governance
- the rules for what may run, who approved it, and the evidence kept.
How it runs, step by step
Before we start
Discovery
- Scored use cases
Analysis
- Systems and data review
- Risk and readiness findings
Readout
- Build, buy or platform memo
- 90-day roadmap
A pilot, defer or stop decision
The steps of an assessment, where your team takes part, and the one point where a person decides.
The steps in words
Kick-off
You: Name a sponsor, a technical contact and a data owner for each workflow. Sign the scope.
Us: Agree the scope and send the list of documents we need.
Discovery
You: Three to five interviews and the documents we asked for.
Us: Interview, read, and draft the scored list of use cases.
Analysis
You: Access to systems and sample data, and answers to our questions.
Us: Audit systems and data, review ownership and approvals, place you on the maturity scale.
Readout
You: The sponsor attends the readout and records a decision for each workflow.
Us: Present the report, the roadmap and our recommendation.
A person decides here. Your sponsor records pilot, defer or stop for each workflow.
How long it takes depends on how many workflows and systems are in scope, so we agree the scope and the schedule with you at kick-off. The optional governance module adds a second readout.
Who a readiness assessment is for
It is for owners and technology leaders who think AI agents could help but cannot yet say where, how safely, or in what order. You do not need a pilot or an AI team. You need a sponsor who can make a decision, and a few workflows you would like to improve.
The questions it answers
Where could an agent take work off my team?
Each candidate workflow is scored, so you can see which are worth doing first.
Are our data and systems ready?
We check access, quality, speed, identity and logging for each workflow.
Who would own it and approve its actions?
We look at ownership, approvals, and what happens when something goes wrong.
Should we build, buy or use a platform?
A short memo compares the options for your situation.
What should the next 90 days look like?
A roadmap with milestones, owners and dependencies.
Which ideas should we start, hold or drop?
A pilot, defer or stop decision for each workflow, with the evidence behind it.
If you are a technology leader: every scored workflow carries its evidence and a named reviewer on your side, and the report states its own limits, including what we did not examine.
What you get
A written report and the working papers behind it: scored use cases, a review of your systems and data, risk and readiness findings, a 90-day roadmap, and a pilot, defer or stop decision for each workflow. You keep everything we produce.
Scored use cases
Each workflow scored for value, feasibility, risk and data readiness, with how much the agent may do without a person.
Systems and data review
What the agent would need to reach, and whether it can: access, data quality, speed, identity and logging.
Risk and readiness findings
Ownership, approvals, incident path and skills, and where you sit on the five-level maturity scale of our AELSHA™ ASTRA framework.
Build, buy or platform memo
A recommendation, with reasons, for the workflows worth pursuing.
90-day roadmap
Milestones, owners and dependencies, plus a cost shape that lists its assumptions and presents none as a measurement.
A pilot, defer or stop decision
Recorded for each workflow, with a date and an accountable owner on your side.
You keep everything
The report, the scoring workbook, the decision log and the interview notes (redacted as agreed) are yours. A named owner on your side takes the roadmap.
An assessment does not build anything, give legal opinions, certify you against a standard, or buy tools for you. An optional governance module can add an inventory of your AI, risk tiers, and a gap register against the rules you name.
The maturity scale, level by level
- L1 Reactive
- L2 Instrumented
- L3 Managed
- L4 Optimized
- L5 Self-improving
The technical view: what the report contains
- Executive summary: one page, with the decision for each workflow.
- Method, scope, evidence base and limits.
- Scored use-case inventory: value, feasibility, risk, data readiness, proposed autonomy tier and decision.
- System and data readiness findings, and organisational readiness.
- Maturity placement by dimension, with the reason for each score.
- Build, buy or platform memo, and the 90-day roadmap.
- Cost shape, labelled as assumptions; risks and open questions.
- Appendices: interview list, document list and scoring rubric.
What we need from you
A decision-maker, a few of your people, and the documents and sample data that show how your work runs today. Sample data can be redacted, and we work under your data-handling rules.
People
A sponsor with authority to take the decision, a technical contact, a data owner for each workflow, and three to five colleagues for interviews.
Workflows
The candidate workflows you have in mind and who runs each of them today.
Access and data
Sample data under your handling rules, and monthly volumes (tickets, documents or queries) so we can shape the cost.
Documents
Architecture diagrams, an inventory of integrations, your identity model, your AI, security and privacy policies, and any commitments to a regulator.
How it runs, step by step
Four steps, with your involvement set out at each. The scope, the schedule, the data-handling terms and an NDA are agreed and signed before anything starts.
- 01
Before we start
Sign the scope, data-handling terms and NDA. You name your sponsor, technical contact and data owners, and we send the document list.
- 02
Discovery
We interview your people, read your architecture and data documents, and agree the scoring criteria with your sponsor.
- 03
Analysis
We audit systems and data for each workflow, review ownership and approvals, and place you on the maturity scale.
- 04
Readout
We deliver the build, buy or platform memo, the roadmap and our recommendation. Your sponsor records a decision for each workflow.
More workflows in scope add time. The optional governance module adds a governance review and a second readout.
After the assessment
The decision is yours. You can start a pilot with us, take the roadmap to your own team or another partner, wait, or stop. Nothing is built during an assessment, so a stop costs you the assessment, not a failed build.
Pilot
We can build and productionise the first workflow, with evaluation, tracing and security designed in from the start.
Defer
The roadmap names what has to change first, who owns it, and what to revisit.
Stop
If the honest answer is no, the report says so and why, and you keep it.
We send a written proposal with a fixed scope and fee, agreed before we start. If you continue with us, we propose crediting the assessment fee toward the next step.
Questions people ask
How long does it take?
It depends on how many workflows and systems are in scope. We agree the scope and the schedule with you in writing at kick-off, before any work starts.
Is it confidential? Do you sign an NDA?
Yes. We sign the scope, data-handling terms and an NDA before the assessment starts. We ask for sample data only under your handling rules, and redacted data is fine.
Is it remote or on site?
Most of the work is interviews, reading documents, and reviewing the systems and data you give us access to, followed by a readout with your sponsor. We work remotely by default. If on-site days help, for interviews or the readout, we agree them and their travel costs in the contract before work starts.
What if the answer is "stop"?
Then that is a useful result, and we will say so plainly. The decision is recorded for each workflow with its evidence, and you keep the report and working papers. Nothing has been built, so you have not committed to a build that would not pay off.
Is this a questionnaire or a score I get back automatically?
No. We interview your people, review your actual systems and data, and present the findings to your sponsor at the readout.
Talk to us about your project
Tell us what you want to automate. We reply by email, arrange a short call, and send a written proposal for the scope.