They hired an employment lawyer, a former HR analytics director, a labor economist, and an operations executive from a manufacturing company.
The team did not score companies on slogans.
It asked concrete questions.
How many complaints were closed without secondary review?
How often did employees leave within six months of reporting managers?
Who could change complaint categories?
Did high-performing executives receive different discipline?
Were promotion criteria specific enough to audit?
Did boards see patterns or only individual legal bills?
One year into the program, the team found a problem at a Horizon portfolio company before it became public.
A regional sales vice president generated exceptional revenue but also an unusual number of resignations among women on his team. No complaint alone proved harassment. The pattern showed retaliation risk.
The board investigated.
Several allegations were substantiated.
The executive was removed before a lawsuit or viral video forced action.
James considered that a greater success than anything that had happened at the gala.
No one outside the company knew.
No one needed to.
"This is what I wanted," he told Sarah after the case closed.
"A quiet firing?"
"A system that notices before somebody needs a camera."
Sarah nodded.
"Less dramatic."
"Much better."
The Dignity Initiative grew alongside the investment work.
James resisted naming it after himself.
He also refused to make the gala video mandatory viewing in every portfolio company.
"I don't want people learning to recognize one famous incident," he said. "I want them learning to recognize the mechanism."
The training used anonymized cases.
A hotel guest who received extra verification because staff said he did not look premium.
A warehouse employee whose complaint became attitude problem in the internal record.
A job candidate described as not polished enough despite higher performance scores.
A nurse assumed to be support staff while a less experienced colleague was treated as the physician.
Participants had to identify the decision point where assumption became action.
James attended the first pilot session silently in the back.
The facilitator asked the group, "What makes bias operational?"
A young manager answered, "When someone else has to obey it."
James wrote the sentence in his notebook.
That evening, he visited his father's old neighborhood in Chicago for the first time in almost a year.
The library where he had learned to code no longer had six slow computers.
It had a renovated technology room with thirty workstations, a fabrication lab, and a small electronics bench funded by several donors including Innovation for Tomorrow.
James deliberately asked that his name not appear on the room.
A librarian recognized him anyway.
"You used to sit over there," she said, pointing toward a corner.
James laughed.
"The computer froze every twenty minutes."
"You kept restarting it."
"I thought that was programming."
She smiled.
A group of middle-school students worked on small robots nearby.
One boy looked over at James.
"Are you the rich guy from the video?"
James closed his eyes.
The librarian laughed.