Provide a reference photo + voice sample, write a script, hit Generate — see that person speak your words. Frontend demo, no backend required.