Transparently.AI reads financial statements for signs of manipulation. Its risk engine scores more than 85,000 publicly listed companies against fourteen clusters of accounting risk signals, and Luca, a forensic accounting assistant, takes an analyst deeper into any individual case.
Forensic accounting is exactly the kind of work that suits this. The signals of manipulation are patterns across many numbers over many periods, comparable across peers, which is tedious and mechanical for a person and fast for a model. The company’s framing is that auditors catch roughly 3 percent of wrongdoing, which is the gap it is selling into.
Portfolio managers, risk teams and auditors are the named buyers. No pricing is published. Two things belong stated: a risk score is a screening signal and not an accusation, and treating one as the other is defamatory as well as wrong; and the headline statistics about how many companies manipulate accounts are the company’s own framing of a contested research area.
