You choose the size of the first step. Not the size of the risk.
The five plans below change how much of the operation comes in at once, not the rule. In any of them the stage price is fixed before work starts, the environment is yours and the dashboard stays open for as long as the system runs.
Stage = 2 to 6 weeks, with a written deliverable and a go-live date
The whole ladder, at a glance.
From left to right the commitment grows, not the promise. Almost nobody enters at the rung they imagined at the start of the conversation. And changing rungs costs nothing beyond telling us.
Spark
Measures the bottleneck and what it costs per year. The fee becomes credit if you move to execution.
Starter
One stage in production: one bottleneck cleared, with scope, date and price fixed before we start.
Growth
The operation's system: automation, integration and AI agents under the same rule and the same dashboard.
Scale
Continuous evolution with a dedicated team: staged roadmap and a predictable technical cost every month.
Enterprise
Several fronts in parallel, with compliance and security requirements designed in from the start.
What is already in the stage price.
None of this is an extra line in the proposal, and none of it depends on the rung you pick. What differs between plans is how much operation comes in at a time, not what comes with it.
Team and stage management
Whoever writes the scope, builds it and follows it into production. You do not pay for meeting hours or management as a separate line.
AI cost within the agreed ceiling
Included up to the ceiling defined in the stage, and measured per query in the dashboard. If usage grows enough to approach that ceiling, you hear it beforehand, not in the invoice.
Deployment and dashboard
Shipping to production in your environment and leaving the system instrumented is part of the stage. Monitoring is not an upgrade: an unmeasured system does not count as delivered.
Common questions before choosing.
Why are the prices not on the page?
Because price is per stage, and the size of a stage depends on how your bottleneck is cut, which is exactly what Spark produces. We share reference ranges in the first conversation, no runaround, so you know whether the order of magnitude fits. What we will not do is publish a closed number with no scope: either it comes in high as protection, or low and turns into change orders halfway. After diagnosis, the first stage price comes out fixed and in writing.
How does Spark work?
It is the way in: we map the process, measure where the operation jams and estimate the annual cost of each bottleneck. You leave with a plan prioritized by impact and with the scope, date and price of the first stage, even if you decide not to continue. If you do continue, the Spark fee becomes credit against execution.
Can I stop after one stage?
You can, with no penalty. Each stage is closed in itself: whatever went live keeps running in your environment, with the code and keys in your name. That is why there is no long contract here: what keeps a client is the next stage making sense, not an exit clause.
Can I switch plans midway?
Yes, and it happens in both directions. The change applies from the next stage, because the current one is already locked in scope, date and price. Moving up usually happens when the first delivery reveals more sitting next to the cleared bottleneck; moving down, when the operation absorbed what it needed and the rest can wait.
Is AI cost billed separately?
No. It is included up to the ceiling defined in the stage, and you see consumption per query in the dashboard. It is not a black box that shows up in the invoice. If usage grows enough to approach that ceiling, it becomes a conversation beforehand, with the number in front of both sides.
Do you do one-off, non-recurring projects?
We do: that is Starter and, for bigger scopes, a sequence of stages with a planned end. Every project start involves an environment setup cost, agreed during Spark, before any development. What we do not do is staff people by the hour with no written deliverable.
How do I know it is worth it before signing?
That is exactly what Spark answers: what the bottleneck costs per year against what it costs to fix. When the math does not work, we say so. Diagnoses have ended with a recommendation to build nothing and fix the process instead. Losing that sale costs less than delivering a system nobody uses.
Not sure which rung is yours?
Almost nobody knows at the start of the conversation, and choosing wrong is expensive. Tell us what is blocking you today: diagnosis returns the bottleneck measured and the rung that makes sense, including when it is smaller than you imagined.