The upside case is real because Innodata sits in a growing bottleneck: high-quality data engineering,
model evaluation, and trusted AI deployment support. I expect several more years of strong growth as frontier-model spending broadens into enterprise and federal programs. But I do not underwrite a pure-software
rerating. The likely outcome is a larger company with somewhat better mix and margins, still carrying a
discount for customer concentration and project-based exposure.