Little is known about how Big Five personality traits manifest in specific conversational behaviors or whether these associations depend on the communication medium. We examined 154 individuals interacting with strangers face-to-face or via video. Using manual annotations, computer vision, and LLMs, we examined actor and partner personality associations with turn-taking dynamics, engagement signalling, and topic development. Higher actor and partner extraversion predicted faster turn-taking and pronounced engagement signalling, whereas actor openness predicted slower turn-taking and reduced engagement signalling. In actors, conscientiousness predicted nodding and in partners it predicted broader topic exploration. However, we found no evidence that these associations differed between face-to-face and video-mediated interaction. Personality thus shapes conversational behavior and these interactional fingerprints appear to survive technological transmission.