




"The Modeler Agent is my thinking partner and my learning partner. I implemented a version dimension plan within an hour; a task that would have normally taken me one to two days of focused effort."
Docker is an open source platform that enables developers to build, deploy, run, update, and manage containerized applications. It helps developers bring their ideas to reality by conquering the complexity of app development.
Prior to Pigment, Docker relied upon Planful for financial planning, but the platform lacked flexibility and capabilities. What that meant ultimately was that much of their work remained stuck in spreadsheets.
When this prompted Docker to look for an alternative, Pigment won the team with its flexibility and ease of use.
But what also convinced them was Pigment’s strong AI capabilities: Docker is proud to be an AI-first organization, and its teams are constantly looking for ways to optimize processes that cost the business time.
Variance analysis and monthly close is, for Docker’s strategic finance teams, one of the most important and time-sensitive parts of their regular monthly operating rhythm.
But it was also an incredibly laborious process: manually reviewing actuals vs budget, drafting commentary for each variance, and then summarizing and packaging narratives for executive review.
It’s nobody’s favourite part of the job, and it was limiting the team’s ability to focus on deeper analysis during close - adding pressure to an already intense cycle.
So they turned to Pigment for help.
Docker now uses the Analyst Agent, in conjunction with the MCP Server to summarize key variances, surface notable trends and provide consistent narratives each month.
BvA data and context live directly in Pigment, which Claude ingests to generate structured first-draft variance narratives in the tone and format the team wants — across P&L, headcount, and cash flow.
Finance team members further apply judgment and audience context to refine the output, eliminating the "blank-page" problem while preserving human ownership of insights. The Analyst Agent surfaces notable trends and flags key movements to support this process.
The results are clear:
Docker also makes use of Pigment’s newest agent: the Modeler.
It acts as a force multiplier, helping the team access more of the platform’s power, faster.
An analyst mentioned that they didn’t have time to look into how Pigment Versions worked. They asked the Modeler Agent to explain the value, and it laid out a plan for them to implement. EPM can be complex, but with AI the barrier to entry for some of Pigment’s most valuable features is so much lower.
The Modeler Agent has also improved the quality and performance of Docker’s models.
Perhaps most importantly, it reduces cognitive load.
Looking ahead, Docker plans to extend the Analyst Agent's role into the annual planning cycle.
By combining the Agent's ability to surface and synthesize historical trends with a custom GPT that the team uses for AOP planning, the team aims to produce a comprehensive NTM (next twelve months) planning playbook — one that brings prior year context, variance patterns, and forward-looking planning questions together in a single, structured starting point.
The goal is to make planning season less about assembling the picture from scratch and more about shaping the strategy it points to.
To learn more about Pigment AI, click here.