About the role
About Cantina: Cantina Labs is a social AI company, developing a suite of advanced real-time models that push the boundaries of expression, personality, and realism. We bring characters to life, transforming how people tell stories, connect, and create. We build and power ecosystems. Cantina, our flagship social AI platform, is just the beginning. If you're excited about the potential AI has to shape human creativity and social interactions, join us in building the future!
About
the Role: As a Prompt Engineer at Cantina Labs, you will operate at the frontier of social generative AI—shaping how millions of users experience storytelling, and connection through AI characters. You’ll collaborate closely with product, design, and engineering to discover what makes AI characters engaging, scalable, and hyper-realistic—and translate those insights into prompt systems, datasets, and evaluation frameworks that power Cantina’s platform. This role bridges research and product, pairing rigorous prompt engineering with strategic reasoning about model behavior and user experience. You’ll own and shape how bots think, speak, and act across the product’s features, balancing immediate product needs with long-term behavioral goals. Your work will directly impact the consistency, expressiveness, and realism of Cantina’s AI characters at scale. What You’ll Do: Define and author core behavior system prompts that shape how Cantina bots express personality, emotion, dialect, and social behavior across text, voice, image, and video. Identify key personality traits that make AI personas compelling, and create datasets and prompt structures that reliably reproduce those traits. Prototype new bot traits and product features through prompt engineering and iterate based on qualitative and quantitative evaluation. Balance short-term feature development with long-term behavioral consistency and quality across the Cantina ecosystem. Collaborate with research to build and maintain behavioral evaluations that measure bot performance, personality fidelity, and interaction quality over time. Partner with product, design, and engineering to translate creative intent and user feedback into scalable prompt systems. Review, refine, and evolve prompts as they move into production, helping establish best practices and launch criteria. Track emerging issues in bot behavior, prioritize improvements, and drive prompt-level solutions in ambiguous or high-impact areas. Document model quirks, prompting strategies, and guidelines to scale prompt development across teams. Support research and ML in our training efforts by delivering meta-prompts for our synthetic data pipelines and evaluating new model behavior What You’ll Bring: 3+ years of experience prompt engineering and developing LLM-based features end-to-end. Strong experience writing, testing, and iterating on complex multi-step and system-level prompts. Deep understanding of LLM capabilities, limitations, and quirks, with strong intuition for how input design shapes model behavior. Experience creating datasets and evaluating model outputs in collaboration with ML or product teams. Strong grounding in psychology, sociology, linguistics, or related fields as they relate to personality, behavior, and dialect. Methodical, hypothesis-driven approach to experimentation and iteration. Ability to manage multiple initiatives, prioritize effectively, and operate independently in ambiguous, fast-moving environments. Excellent communication skills, with the ability to explain prompt design decisions to technical and non-technical stakeholders. Comfortable communicating with executive-level stakeholders. Basic Python familiarity and comfort navigating unfamiliar codebases