The next round of Remix agent hackathon is underway, and this time, Kinaxis partners are taking the controls.
Cohort 1 showed what customers could build from the supply chain challenges they know firsthand. Across 18 teams, participants turned their own planning realities into agent concepts built around practical business problems.
Now, 15 partners have entered the two-week hacking period for cohort 2, running August 17-28, bringing experience from across customer organizations, past implementations, and industries into Maestro Agent Studio.
That shift introduces a different starting point. Customers bring deep knowledge of how planning happens within their own organizations. Partners can draw on patterns they have seen across many.
The Partner Cohort is officially in build mode.
Why the partner perspective matters
While every supply chain operates differently, similar friction points can surface across planning environments.
Planners may spend hours investigating disruptions, gathering context from several systems, or comparing responses before making a decision. Through implementations, industry experience, and customer conversations, partners see where these challenges repeatedly slow down planning work — and where different organizations encounter variations of the same underlying problem..
That experience gives them a different lens for identifying agent opportunities. The question is not simply what problem exists in one planning environment, but which challenges keep appearing, why they persist and where a focused agent could make that work easier.
Remix gives them two weeks to turn that pattern recognition into something testable.
Inside the two-week build
Teams are moving from broad observations to focused use cases, narrowing the planning problems they want to address and defining exactly what their agents should accomplish.
From there, they build, test, and refine their concepts in Maestro Agent Studio. For partners, that means translating experience into design choices:
- Which planning challenges have surfaced repeatedly?
- What part of that challenge is common enough to recognize, but focused enough for an agent to address?
- How can that accumulated expertise be turned into a practical workflow for the planner?
- Each decision moves the concept from something partners have observed in the field toward something they can put to the test.
The short sprint rewards focus. Teams do not need to solve every part of a planning process. They need one simple, functional concept that shows why their use case matters: Find a meaningful planning problem. Give the agent a clear job. Build enough to show its value.
Start with the problem, not the AI.
The bigger question behind the build
Each team is focused on one agent concept, but Remix cohort 2 is exploring something broader: What happens when expertise built across many supply chain transformations can be translated into something an agent can apply?
Partners bring more than familiarity with individual planning challenges. They bring perspective on how those challenges show up differently across organizations, which approaches tend to work and where planners consistently need more support.
Agents create an opportunity to turn some of that expertise into workflows that can investigate, interpret, and help users act.
Cohort 2 gives partners a place to test what that could look like in practice. The outcomes are still unknown. And that is exactly what the build is designed to explore.
The next phase of Remix
When the sprint ends, the interest will not only be in what partners built. It will also be in why they chose those problems which patterns they recognized and how their experience across planning environments shaped each agent's role.
After the Partner Cohort concludes, the next Customer Cohort will bring another set of planning challenges and perspectives into Remix.
For now, the concepts are still taking shape. Andrew Bell, Chief Product Officer at Kinaxis and one of the Remix cohort 1 judges, captures the anticipation simply: “I can’t wait to see what they’re going to build.”
The next useful agent may not start with a new problem at al. It may start with one partners have seen too many times — and a new way to solve it.
