Buying an AI tool is easy. Getting an organization to actually use it, in ways that produce real value, is a different problem entirely.
Designed and led structured AI pilot programs with defined kickoffs, custom workflow development, and feedback loops built around what the organization is trying to accomplish.
15 organizations taken through structured pilots, with workflows built from real feedback and enterprise engagements that continue today.
Research consistently shows that somewhere between 70 and 85 percent of AI deployments fail to deliver the value organizations expected. The reason is almost never the technology. It’s the lack of structure around how the organization actually adopts it.
Most AI pilots look the same: a vendor gives a demo, a few people try it for a week, and then it quietly gets shelved because nobody knew what they were trying to accomplish in the first place. There was no definition of success, no process for collecting real feedback, and no one building the specific workflows that would have made the tool useful for that particular organization.
That’s the problem I work on.
A real AI pilot starts before anyone touches the tool. The first question is not “how does this work” but “what are we trying to get out of this, and how will we know if we got it.”
That means a kickoff where you understand the organization: what they do, what they struggle with, what workflows would actually save them time or surface insights they don’t have today. Then you build those workflows. Not generic templates, but ones configured around what that organization is actually trying to accomplish.
From there you need a feedback loop. What is getting used? What isn’t? What are people asking for that doesn’t exist yet? The answers to those questions drive what gets built next.
I have designed and led AI pilot programs across two contexts.
At an enterprise software company, I worked directly with nonprofits piloting an AI tool built on their own documents and organizational data. Each organization went through a structured kickoff. I set up their environment, configured the platform around their specific needs, and built out a library of workflows covering everything from grant proposal writing and impact reporting to email campaigns and social media content. Every workflow was built to produce outputs that were specific to that organization, not boilerplate.
Feedback from those pilots drove the workflow library. When an organization came back and said they needed something that didn’t exist yet, that became the next thing to build.
That work continues through engagements like a pilot program with an enterprise AI platform for funders, working through how the tool fits into their existing processes and building workflows around their actual grant documentation and organizational knowledge.
Most AI tools are capable of more than the organizations using them realize. The gap is not in the technology. It’s in whether anyone took the time to figure out what the organization actually needed and built toward it.
That is the work. And it is ongoing, because a good pilot does not end at go-live. It ends when the organization no longer needs someone to tell them what to do with it.
Have a similar problem in your business?
Get in Touch