AI Prompt & Agent Developer
About the role
Who Are We? A fast-growing, seed-stage AI healthcare infrastructure company is building voice and text operations software for medical practices at serious scale. The company’s AI agents already support approximately 1,000 physical healthcare locations and process tens of millions of calls, helping practices automate front-desk workflows, patient communication, scheduling, intake, insurance verification, and other high-volume operational tasks. This is not a generic chatbot company. The team is building production AI agents that must perform in live, messy, high-stakes customer interactions where latency, language nuance, call flow, tone, accuracy, and reliability all matter. The product sits at the intersection of voice AI, healthcare operations, prompt engineering, agent behavior design, evaluation systems, data curation, and production-quality automation. Backed by a premier startup accelerator and already operating at meaningful customer scale, the company is now expanding its AI engineering team in San Francisco. You’ll join a small, intense, highly execution-oriented group building systems that move from real call data to production improvements daily. This is a rare opportunity for someone who loves language, data, systems, and customer behavior to own the production behavior of AI agents used in the real world. What’s in It for You? Direct ownership of production AI behavior: Own behavioral slices of voice agents that interact with real patients and healthcare practices every day. Massive real-world data loop: Work from live call data, customer feedback, dashboards, evals, and measurable outcomes instead of abstract model benchmarks. High-scale AI product exposure: Build for AI systems already used across roughly 1,000 physical locations and tens of millions of calls. Fast iteration environment: Review failed calls, identify patterns, ship prompt and agent improvements, and measure impact in tight feedback cycles. Rare blend of language and engineering: Apply writing sensibility, analytical rigor, prompt architecture, and practical coding to improve production AI agents. Career acceleration at an early-stage company: Join a YC-backed seed company as it scales from millions of calls toward much larger enterprise volume. Healthcare impact: Improve front-desk operations for medical practices, helping patients get scheduled, routed, verified, and supported more efficiently. What Will You Do? Prompt & Agent Behavior Ownership Write, maintain, and improve the prompts that run in production across live healthcare voice agents. Own key behavioral flows including intent classification, information extraction, availability negotiation, closing phrases, insurance verification, objection handling, and edge-case recovery. Design shared behavior that applies across deployments as well as customer-specific behavior for individual healthcare practices. Shape the tone, register, rhythm, and wording of voice agents so they sound natural, helpful, and contextually appropriate in American English. Data Review, QA & Iteration Analyze real failed calls, identify recurring breakdowns, and turn call-quality insights into prompt and agent improvements. Simulate real-world customer scenarios to test whether agents respond accurately, naturally, and safely. Monitor production agent behavior to catch drift, regressions, and quality issues early. Push automation rates upward through small, measurable, data-backed wins. Evaluation & Optimization Build offline evaluation sets that let the team test prompt and agent changes before they affect live deployments. Create evaluation harnesses, QA workflows, and test suites for production AI behavior. Run automated prompt optimization workflows using GEPA-style approaches and related methods. Establish metrics for customer-specific deployments, such as booking rate, successful routing, insurance verification completion, recovery success, and call resolution quality. Customer Onboarding & AI Configuration Work on human-in-the-loop onboarding systems that translate customer intake forms, scheduling rules, operational quirks, and workflows into agent configurations. Help improve internal AI tools that accelerate deployment setup and ongoing configuration quality. Collaborate across engineering and customer-facing teams to understand the messy details of healthcare practice operations. What Will You Need? 2+ years of experience with AI/ML, NLP, prompt engineering, context engineering, data curation, or production evaluation workflows. Hands-on experience owning AI product behavior in production, ideally for systems used by 1,000+ users. Strong prompt engineering experience with real user feedback, production failures, measurable iteration, and ongoing quality ownership. Experience with data curation, call review, evaluation design, quality review, or production AI monitoring. Excellent written and verbal communication skills in American English. Strong sensitivity to language, tone, register, phrasing, cultural norms, and how wording changes user experience. Comfort analyzing data for long stretches, spotting patterns, and making disciplined, empirical improvements. Ability to justify product changes with metrics rather than intuition alone. Working knowledge of Python and comfort reading TypeScript; stronger Python skills with APIs, data pipelines, or testing frameworks are a plus. Comfort using modern AI coding tools while still understanding what the code is doing. Strong analytical and problem-solving ability in ambiguous environments. Bachelor’s degree or equivalent experience in a relevant field such as Computer Science, Engineering, Math, Philosophy, Linguistics, Cognitive Science, English, Medicine, or a related discipline. Bonus: experience with voice AI, TTS, ASR, telephony platforms, Twilio, contact centers, SaaS, customer-facing AI, healthcare operations, dental or medical scheduling, automated prompt optimization, DSPy, GEPA, MIPROv2, or fine-tuning. Bonus: for non-native English speakers, a linguistics, translation, or equivalent background plus strong familiarity with American cultural norms and U.S. customer-facing communication.
What we are looking for
- 2+ years in prompt development, data curation, evals in production
- Strong analytical and problem-solving mindset; comfort with ambiguity
- Excellent written and verbal communication skills (U.S. English)
- Can Vibe code - Python chops beyond reading: APIs, data pipelines, testing frameworks
- Prior work with voice AI, TTS, ASR, or telephony platforms (Twilio, etc.)
- Contact center, SaaS, or customer-facing tech background
- Healthcare or medical operations experience, preferred
- Automated prompt optimization experience (DSPy, GEPA, MIPROv2)
- Bachelor's degree and/or extensive experience in one or more of: Computer Science, Engineering, Math, Philosophy, Linguistics, Cognitive Science, English, Medicine, or a related field