How it works
From roadmap to release, in one system.
People set direction and make the calls. AI agents do the heavy lifting. Every hand-off between them is explicit, and every step is on the record.
Start from what the organisation intends to build.
Every product keeps a ranked roadmap. Open a cycle from any item, and the roadmap keeps track of every cycle it spawned and what each one produced.
A plan, sized and assigned — approved by a person.
A planning agent breaks the work into items, each with a clear definition of done and the right agent to do it. Nothing runs until someone on your side approves the plan. Approving it is what creates the work.
Every task starts with the right context.
Before an agent begins, Wakalix assembles its context package: what's being asked, the parts of the codebase that matter, your standards and rules, and the decisions already made. The agent starts informed — not cold.
Agents build it — and stop to ask when the call is yours.
A developer, an architect, a reviewer, a product manager — each with a defined job and hand-off, running in your engineers' own harness on the subscriptions you already have. When an agent reaches a decision that belongs to a person, it stops, lays out the options and waits. It doesn't guess.
Checked against its criteria. Accepted by a person.
Every piece of work comes back with its evidence: the commit, the branch, the agent and version that produced it, and a verdict against each acceptance criterion. A person accepts it, challenges it or sends it back. It lands as a pull request in your own repository — and a person merges it.
Every cycle makes the next one better.
At the end of a cycle, the team looks back. What an agent refused, got wrong or had to ask about goes back into the agents themselves — so the next cycle starts from a better team, not from scratch. And you can see how the quarter's work moved: what was planned, what landed, and where it waited.
