Training

Training for technical teams using AI

Training is not a tour of tools. We work on tasks similar to real repository work: code discovery, planning, implementation, tests, review and context handling.

Discuss training

Scope

What we review and what result the change should deliver.

The signals help identify the problem. Scope and result show what RUNSY can own technically.

Signals

What points to the problem

the team tests AI tools, but everyone works differently and quality is hard to judge
AI helps with isolated tasks, but planning, implementation and review rules are missing
the company wants to use AI in software work without lowering technical standards

Scope

What we change

adapt the workshop to the team's repositories, roles and current problems
practice code discovery, change planning, implementation, tests, review and decision records
define control rules for when an agent can work independently and when review is required

Result

What should work differently

a more repeatable way to work with coding agents and AI tools
fewer chaotic experiments and more changes that can be reviewed in the repository
better use of AI without losing control over architecture, tests and security

Related situations

Example problems from this area.

See how this area connects with specific problems in processes, data and systems.

Training

The team uses AI without shared work rules

Problem

People use coding agents and AI tools in different ways, so good results are hard to repeat and changes are hard to review.

What we change

We work on tasks close to the team's actual work: code discovery, planning, implementation, tests and review. We also define when an agent requires human review.

After the change

The team has shared rules for using AI and can verify the result in code, tests and documentation.

Training

See all examples

Let's talk

Start with the process that takes too much time today.

Tell us where the team loses time, which systems are involved and what needs to work differently. That is enough for a first conversation.

Contact

Describe the problem, and we will start with a short diagnosis.

Send a few sentences about the process, systems and expected result. That is enough to see whether the first step is integration, AI automation, an audit or an operational application.