ClearLedger Analytics

ClearLedger Analytics is a regional risk-modeling outfit operating out of a converted bank vault on the third floor of an Old Pittsburgh office building that was, until a few years ago, an actual bank. The vault remains, repurposed as the server room. Five analysts and a founder. Their main product is anomaly-flagging models for small and mid-size credit unions; their main competition is two larger firms in the same downtown core and an offshore dev shop they have never met but suspect undercuts them on every bid.

They invoice in five-figure Cu, net seven days. They request models with seven-day deadlines. There is a culture here.

The founder describes the company on her professional profile as “the place where the work actually gets done,” which her analysts find both flattering and ominous. She has been overheard, at industry events, describing the AI dev sector as “the gold rush, but with thirty-day net terms.”

Their server room — the converted vault — contains, beside the actual production cluster, a second small rack that is never explained on tours. Industry rumor (likely unfounded) holds that ClearLedger keeps one experimental personal-research model running on it, off the books. The founder denies it. The analysts roll their eyes.