AI Research Engineer

As an AI Research Engineer at Amperos Health, you'll get to work on problems at the intersection of healthcare and the frontier of applied AI. From overcoming anti-AI payors to improving the quality of results collected from an insurance agent, you'll pave the latest developments in healthcare AI. About the role As an AI Engineer on the team, you will be focused on passing the healthcare billing Turing Test - would an insurance agent who is interacting with our AI or a healthcare biller who is reviewing the work of our AI think our AI is a human, and not an AI? More concretely, you will be working to improve our AI agent's quality so it acts more human-like and collects more accurate and detailed results from insurance. Success for the role during the first 3 months looks like implementing several potential solutions, accompanied with measurable evaluations, that lead to multiple insurances that were previously anti-AI being unlocked and a substantial improvement in the quality of claims info collected. Responsibilities Explore fine tuning, prompt engineering, and agentic framework solutions Lead efforts to improve our AI system's performance against anti-AI detection measures (ie. improving voice quality, implementing interruption models, adding more human-like speech patterns) Apply the latest multi-agent framework research to improve AI calling about claims Develop and implement quantitative metrics to evaluate system realism and performance Optimize infrastructure for reduced latency through self-hosting and model optimization Enhance conversational flow by addressing interruptions and question thoroughness Improve call quality by developing detection systems for various scenarios