The Role
Cedar is seeking an AI Engineer/Applied Scientist III to join our Support/Kora squad. In this role, you will design, build, and optimize Kora, an AI Voice Agent, purpose-built for patient billing and the central product in Cedar’s Support offering.
Your day-to-day work centers on prompt engineering, conversation flow design, and context optimization, not training models from scratch. You’ll use foundation model APIs and agent SDK platforms to craft AI behaviors that help patients understand and resolve their bills over the phone. You’ll combine rigorous data analysis (e.g., reviewing call transcripts and patient account data) with production-level coding to ship improvements that make a measurable difference for real patients and healthcare providers.
Responsibilities
- Design and optimize Kora’s behavior, including prompt engineering, context window design, and conversation flow architecture, to help patients understand and resolve billing issues over the phone
- Implement your solutions in production code (Python and TypeScript), leveraging AI coding assistants as part of your toolkit
- Design robust experiment plans that deconstruct large goals into small testable hypotheses. Execute these plans through iterative development cycles where you ship new features, evaluate how your features work with real patients, and continue to improve and optimize Kora’s behavior, displaying a strong ownership mentality, product mindset, and user experience focus
- Analyze unstructured data, call transcripts, patient account details, conversation logs, using both quantitative and qualitative approaches to identify patterns and improve agent performance
- Partner closely with other scientists, engineers, designers, and product counterparts to solve problems as a team
Skills and Experience
- 3 or more years of full-time experience working in a structured product development environment on a product that had frequent releases
- Hands-on experience with LLM APIs (e.g., OpenAI, Anthropic, Google) and/or AI agent frameworks — including prompt engineering, context management, and evaluation
- Experience with modern programming languages (in this role, you’ll use both Python and TypeScript), as well as experience with AI-enabled coding
- A strong understanding of both the scientific method and software engineering principles, with a proven ability to balance experimental rigor and rapid production deployment
- Experience collaborating with partners from different job functions, and the ability to translate business requirements into technical builds
- Ability to thrive in an entrepreneurial environment, showing a tendency toward speed and action while achieving clarity from ambiguity
Nice to Have:
- Experience with conversational AI voice agents
- A track record of transforming raw, unstructured data sources into actionable intelligence
- Experience in healthcare, healthtech, or other regulated domains