Agentic engineering for your developers
Has your in-house development team made the switch to agentic engineering yet?
There is a difference between using AI to help you write code and a full switch to agentic engineering, and that switch, done well, could get you there for less effort, and you’ll see the results faster.
What is agentic engineering?
Agentic engineering is a way of building software in which AI agents write, test and change the code, and your developers decide what to build, direct the agents and check their work.
With an AI assistant, the developer writes the code and the AI suggests lines. With an agent, the developer describes the change, the agent makes it and runs the tests, and the developer reviews it.
Do you work this way yourselves?
Yes. We built The GAiGE, our AI governance and adoption platform, with agentic engineering. It’s live, so you can see the result for yourself.
How does a team make the switch?
It works best with support, while your developers get used to directing agents on real work.
A session on your own code
We work with your developers on a real task from your backlog, using agents on your own codebase. You see what it changes before you decide on the next step.
A way of working for the team
How each piece of work is written up before an agent starts, how big each change should be, and how it’s checked before it’s released.
One team at a time
The first team makes the switch on its everyday work, with us alongside and the tools set up. Other teams follow once it’s working this way.
Would you rather we built it for you?See Custom software
What has to be in place before agents change your code?
We put these checks in place with your developers before an agent starts work:
- Automated tests on every change
- A developer reviewing each change before it’s released
- Clear rules for passwords, keys and customer data
- Limits on what an agent is allowed to touch
What else would you like to know?
Is code written by AI agents secure?
It can be, with the right checks around it. Code from an AI agent can run perfectly well and still have security holes nobody has looked for.
So the switch includes those checks: automated tests on every change, a developer reviewing each one, and clear rules for passwords, keys and customer data.
Does this mean we’ll need fewer developers?
It depends on what you do with the time it frees up, and that’s your decision. The switch changes what your developers spend their time on: much less on routine code, and more on the decisions, the design and the checking that only people can do.
How much faster will our developers be?
When we develop software, we’re seeing development time cut by up to 80%, though every project is different. For your team, it depends on your code and the work you do, which is why the first session is on your own code.
When is the switch worth making?
It’s worth making when your developers already use AI, mostly for suggestions and answers; when there’s more work waiting than your team can get through; and when the software is your own, and you’ll keep changing it for years. If your developers rarely change the software, the switch can wait.
Do you need access to our code?
Yes, for the session on your own code. We work within the access you give us and your own rules, and we’ll sign a confidentiality agreement before we start, if you’d like one.
How much does it cost?
We always like to start with something manageable, and because we use AI in everything we do, you’ll probably find we cost a lot less than you might expect. Let’s have a brief meeting about your specific requirements, and then we can let you know the costs.
What else could help?
Is it time to make the switch?
If you think your developers are ready for agentic engineering, chat to Aimee and she’ll help you set up a first meeting.