In our organization we have finance validate - all the savings/cost avoidance - sourcing reports to maintain trust and integrity in our numbers. We have enabled workflows and a technology driven process to enable this task. Do any of you do it another way while still maintaining the integrity of your numbers? Do any of you establish thresholds below which the validations do not need to happen or can be done by stakeholders? How do you maintain consistency in how different validators approach the numbers? Do any of you have a centralized validation approach? How did you get there?

Our saving & Cost avoidance are also validating by our Finance partner to ensure integrity of our numbers. We do not have any thresholds, all projects are approved by our Finance partner. To maintain consistency, we have established a policy which defines precisely the saving calculation rules for every type of saving. We also have established templates for each of them, so all saving projects are provided to finance using the same template & calculation methodology.
The validation is centralized and done by our Finance partner (1 person). The saving policy & templates were done in collaboration between Procurement & Finance to ensure full alignment & complete understanding.

The approach is similar, as validators are within finance region, then to finance group, however, specific documentation-evidence is used to demonstrate the values claimed. I question how this can be mechanized though?

Past experience: we agreed on 4 ways to calculate the savings (agreement between finance and procurement/sourcing, at group level), so the savings had to be calculated/validated using that methodology and nothing else. It was pretty straight forward and there were no back doors (you could not match one saving to multiple types, but only to one).