The River Group

Work with us

Who this is for, and who it is not

Four things have to be true. If they are not, this is not the work for you, and finding that out here beats finding it out three conversations from now.

  1. You own an outcome with a number on it, and you are measured on it. Not a budget, and not a mandate. A number on a scorecard: cycle time, cost to serve, retention, loss ratio, first-contact resolution, days to close. If the operation behind the number isn't running yet, the target you will be measured on counts. If you have been asked for an AI strategy, that number is where it starts, because there is only an operating strategy for your organization, and a position on how it uses agentic AI to execute that strategy.

  2. Being wrong is expensive. Changing direction in your scope may take more than a quarter, may cross more than one function, and cannot be quietly reversed. Where a position can be abandoned in a week at no real cost, there is nothing to deliberate: try it and correct. Where being wrong means eight months and a reorganization, the position is worth taking on purpose.

  3. At least one outcome you own runs through an operation of a particular kind. One where the volume is high and not going down, the steps repeat, the information needed is scattered across several systems, and most of the day goes on reading things, looking things up and applying rules, with judgment saved for the exceptions. Agents are changing what that work costs now, whether you engage with it or not. Your competitors are finding out at the same time you are.

  4. Something is forcing it now. A board question, a budget cycle, a competitor's move, a pilot that ran and left nothing behind, a number that has stopped responding to the things that used to move it. Without one of these, the topic is interesting and nothing has to happen.

Where it is not a fit. If nothing you own has a number on it, or a target you will be measured on, there is no position to take. If the operations you run are judgment end to end, low volume and non-repeating, agents will assist them over the next 12 months and will not change their economics, and a conversation about how far to reimagine them would be a conversation about very little. And if what you want is someone to build the thing, we are the wrong call.