Highly Experienced Prompt Engineer

We're seeking a creative and highly analytical Prompt Engineer to join our dynamic team and play a pivotal role in shaping the core of our AI simulation technology. You'll be instrumental in designing the interactions that rigorously test and evaluate AI voice agents, ensuring our platform effectively pushes the boundaries of AI capabilities and uncovers potential weaknesses. Your work will directly contribute to the trustworthiness and performance of our clients' voice AI systems. What You'll Do - Design & Develop Prompts: Craft, refine, and manage extensive libraries of prompts designed to test the capabilities, robustness, safety, and conversational fluency of diverse AI voice agents across various scenarios (e.g., customer support, booking, ordering). - Scenario Simulation: Develop and implement sophisticated prompt strategies that simulate realistic user interactions, edge cases, adversarial inputs, and complex conversational flows using our platform's workflow and persona features. - Analyze & Evaluate: Systematically analyze the responses and conversational behavior generated by AI voice agents in response to test prompts, identifying failure modes, inconsistencies, and areas for improvement. - Collaborate: Work closely with AI researchers and software engineers to integrate advanced prompting techniques and evaluation methodologies into the Cekura testing and observability platform. - Stay Current: Keep abreast of the latest advancements in Large Language Models (LLMs), conversational AI, voice synthesis/recognition, and cutting-edge prompt engineering techniques relevant to voice interactions. - Iterate & Improve: Continuously experiment with prompt variations, analyze their effectiveness, and iterate on prompt design based on testing results, emerging AI capabilities, and client feedback. - Data Integration: Contribute to strategies for leveraging past conversations and real-world call data to generate more effective and realistic test prompts and scenarios. - Document: Maintain clear and organized documentation of prompt libraries, experimental designs, evaluation findings, and best practices.

Cekura (formerly Vocera) is a Y Combinator–backed (F24) startup revolutionizing the testing and observability of AI voice agents. Founded by a team with deep technical expertise, we address the critical challenge of ensuring voice AI reliability in a market flooded with new solutions. Manually testing voice agents is slow and often fails to catch crucial edge cases before production. Cekura automates this process, simulating thousands of realistic conversational scenarios using sophisticated workflows, dynamic personas, and real-world data. Our platform empowers companies building voice agents for customer service, sales, appointments, and more to test efficiently, identify issues early, demonstrate reliability, and deploy with confidence. We make AI voice agents dependable. Learn more at cekura.ai.