"How much does this cost" is usually the first question, and it's the one we can answer least honestly with a single number, because two projects that sound identical in a first conversation can differ in cost by five times once we actually scope them. The tools rarely explain that gap. These five things usually do.
1. How many systems the process actually touches
A process that lives entirely inside one platform is a fundamentally different project from one that spans four systems with no shared source of truth. Every additional system is another integration to build, another set of edge cases, another thing that can change underneath you later. This is the single biggest driver we see, and it's usually invisible until someone actually maps the process step by step.
2. How clean the existing data is
Automating a process that depends on well-structured, consistent data is comparatively quick. Automating the same logical process on top of years of inconsistent naming, duplicate records and undocumented exceptions means a meaningful chunk of the project is data cleanup before the automation itself even starts. This is the cost driver most often underestimated on both sides of the table.
3. How much judgment the process requires
A stable, rule-based process is a workflow automation or RPA project, and those are relatively fast to scope and build. A process with real ambiguity — where the right action depends on context a rule can't fully capture — needs the more careful design work a custom AI agent requires: defined guardrails, evaluation, a real handoff path. More judgment generally means more cost, not because the AI itself is expensive, but because getting it trustworthy takes real design time.
4. How much human safety net the launch needs
A system that can launch with a human reviewing every action is faster and cheaper to ship than one that needs to run unsupervised from day one, because the guardrail work is lighter when a person is still the final check. Most of our engagements start supervised and earn more autonomy over time — which is also usually the more cost-effective path, not just the more cautious one.
5. Whether this is a one-time build or a system that needs to evolve
Some automations are genuinely "build it once, it keeps running." Others sit inside a process that changes every quarter, which means the real cost includes the ongoing tending, not just the initial build. Knowing which kind of project you're actually commissioning changes the right way to scope and budget for it.
The honest answer
We don't publish a price list, because a number that ignores these five factors isn't useful to you — it's either scary-high to cover the worst case, or unrealistically low to win the conversation. What we do instead is scope the actual process first, usually in a short strategy conversation, and come back with a number that reflects what your specific project actually involves.