Most of the upside comes from revenue compounding, not heroic
rerating. Innodata already has real AI demand, good margins, net cash, and low capital intensity, so it can scale faster than traditional IT services peers. I give it some mix improvement if evaluation,
observability, and assurance become sticky, but I do not underwrite a pure-software valuation because customer concentration and service exposure remain material.