Contents
An AI readiness check shows you quickly whether your development team is ready to use AI coding tools and coding agents productively – and where the biggest gaps are. With the following 24 questions across six dimensions you can assess your team yourself. Depending on your score, you get concrete next steps at the end.
Why an AI Readiness Check Before Rollout Makes Sense
In software development AI acts as an amplifier: good processes get faster, weak processes produce problems faster. Teams that start without taking stock often invest in licences while the real bottlenecks lie elsewhere – missing tests, unclear data protection rules or limited experience in the team.
A readiness check answers three questions: Where do we stand? What holds us back most? Where do we start? It does not replace a detailed analysis, but it creates a shared basis for discussion between management, team leads and developers.
How the Self-Assessment Works
Answer each question with "yes" (2 points), "partly" (1 point) or "no" (0 points). Each dimension has a maximum of 8 points, 48 in total. The result is most meaningful when several people – for example the CTO, a team lead and two developers – answer independently and you compare the results afterwards. Large discrepancies are an important finding in themselves.
The Six Dimensions of the AI Readiness Check
1. Strategy and goals
- Have we defined concrete, measurable goals for using AI in development?
- Is there a person responsible for the topic?
- Is budget allocated for tools and training?
- Has management been briefed on realistic expectations?
2. Technical foundations
- Do our most important projects have an automated test suite with meaningful coverage?
- Does every merge pass through a CI pipeline with build, tests and static analysis?
- Are architecture, conventions and setup documented well enough for a newcomer to understand?
- Can projects be set up and tested locally with just a few commands?
3. Data protection, security and legal
- Have we clarified which AI tools are approved and where data is processed?
- Are there rules on which data (customer data, secrets, contracts) must never enter AI tools?
- Have secrets been removed from the code and managed centrally?
- Have we addressed open-source licences and copyright for AI-generated code?
4. Team skills
- Does the majority of the team regularly use AI tools beyond simple autocomplete?
- Does the team know the basics of prompt and context engineering?
- Is there experience with agentic tools that edit files and run tests on their own?
- Have employees received training on the opportunities and risks of AI?
5. Processes and quality assurance
- Are code reviews mandatory for every change?
- Are tasks cut small enough to be implemented and reviewed quickly?
- Are requirements written with clear acceptance criteria?
- Do we measure flow metrics such as lead time or defect rate?
6. Culture and readiness for change
- Is the team allowed to spend time on experiments without delivering full sprint output immediately?
- Are AI experiences shared in the team, for example in demos or knowledge sessions?
- Are concerns discussed openly rather than brushed aside?
- Are there internal pioneers who can support others?
A note on the skills dimension: since 2 February 2025 the EU AI Act has required organisations deploying AI systems to take measures to ensure sufficient AI literacy among their staff (Art. 4). What this means in practice is explained in AI literacy under the EU AI Act – this is not legal advice.
Scoring: Which Maturity Level Fits Your Team?
| Points | Maturity level | Typical picture | Recommended focus |
|---|---|---|---|
| 0–16 | Starting out | Isolated, unregulated use; foundations missing | Create rules, goals and a technical base |
| 17–32 | Building up | Tools in place, usage inconsistent | Training, shared workflows, measurement |
| 33–48 | Scaling | Solid base, first successes visible | Agentic workflows, automation, rollout to more teams |
Besides the total score, look at the weakest dimension. A team with 36 points but only 2 points in data protection and security should start exactly there before rolling out agentic tools broadly.
Next Steps by Result
Starting out
- Appoint an owner and define two or three concrete goals.
- Choose approved tools – criteria are covered in choosing AI coding tools.
- Write a short policy on data, secrets and review duties.
- Backfill tests and CI in a pilot project.
Building up
- Run a joint training on your own code so everyone works at the same level.
- Create project rules and context files for AI tools.
- Capture baseline metrics and compare them after a few weeks.
Scaling
- Establish agentic workflows for recurring tasks such as tests, migrations or documentation.
- Document successful patterns and transfer them to other teams.
- Follow a structured rollout plan such as the 90-day plan for agentic coding.
From Readiness Check to Workshop
The result of the check is an ideal starting point for training: instead of generic content, you can work specifically on the weakest dimensions. We place AI developers from real-world practice as trainers and agree content, format (in-house or remote) and scope individually – workshops start from 799 EUR net. What to clarify beforehand is described in preparing for an agentic coding workshop.
Planning an app? Request a free fixed-price quote in 2 minutesConclusion
An AI readiness check creates clarity before money and motivation flow in the wrong direction. The six dimensions – strategy, technical foundations, data protection and security, skills, processes and culture – show where your team stands and which lever has the greatest impact. Repeat the check after a few months to make progress visible.
Frequently Asked Questions
How long does an AI readiness check take?
Each person can answer the 24 questions in about half an hour. Comparing results as a team takes an additional one to two hours.
Who should complete the readiness check?
Ideally several roles independently, such as the CTO, team lead and developers. Differing assessments reveal where perception and reality diverge.
What should a team in the starting phase do?
Focus first on clear rules, goals and technical foundations such as tests and CI. A broad rollout of agentic tools is premature at this stage.
How often should the check be repeated?
Repeating it after three to six months makes sense to measure progress and set the next priorities.