Applied AI Engineer

Ascertain is building AI agents to automate the administrative work that burdens care teams. We are in major health systems and large specialty groups, saving hundreds of staff hours every week.

Our backers include Northwell Health, New York’s largest health system, and Deerfield Management, a leading healthcare investment firm.

Together, we’re on a mission to restore time, trust, and focus to the people who keep healthcare running. Our work is urgent — not because of startup timelines, but because our customers rely on us to drive financial resilience and operational clarity in a system under strain.

You'll be one of the first dedicated AI engineers focused on agentic systems at Ascertain. Your near-term mandate is concrete: build voice agents that call payers on our behalf to check the status of prior authorization submissions, follow up on outstanding requests, and capture structured outcomes back into our systems. Today, this work is done by humans sitting on hold; you'll replace large parts of that with agents that are faster, more reliable, and never get tired.

Over time, the scope of this role will grow well beyond voice. As we lean further into AI across the company, you'll have a hand in shaping how Ascertain builds, evaluates, and operates agents of all kinds—voice, chat, document understanding, internal tooling, and beyond. We're looking for an AI engineering generalist who's excited to start with voice and build outward from there.

• →

  Design, build, and ship voice agents that conduct outbound calls to payers - handling IVR navigation, hold queues, multi-turn dialogue with human reps, and structured data capture.

• →

  Integrate telephony, speech-to-text, text-to-speech, and LLM components into reliable end-to-end pipelines, and own them in production.

• →

  Build the evaluation and QA infrastructure that lets us measure agent quality, catch regressions, and continuously improve performance against real-world calls.

• →

  Develop and maintain the APIs and integrations that connect agent outputs to the rest of our prior authorization workflow.

• →

  Instrument everything: call outcomes, agent behavior, latency, failure modes - and use that data to make the next version better.

• →

  Partner closely with our operations and clinical teams to identify the highest-leverage workflows to automate next, and translate those into agent designs.

• →

  As the AI surface area at Ascertain grows, expand into other agent and LLM-driven systems across the product and internal tooling.

• 4+ years of professional software engineering experience owning production systems end-to-end—from design through deployment, monitoring, and iteration.

• 2+ years of professional Python experience building APIs, services, or data workflows.

• Hands-on experience developing or deploying LLM-based applications or RAG systems in a real production setting (not just prototypes).

• Comfort operating with ambiguity and shipping the first version of something that doesn't exist yet.

• Hands-on experience with voice or conversational systems - IVR, call automation, speech-to-text, telephony APIs (Twilio, etc.), or conversational agents.

• Experience implementing QA processes and evaluation frameworks for agentic or LLM-based systems.

• Track record of building and maintaining RESTful APIs and third-party integrations.

• Proficiency with observability and monitoring tools for production reliability.

• Experience with version control and automated CI/CD deployment pipelines.

• Comfort managing SQL or NoSQL databases for storing and analyzing interaction data.

• Healthcare, payer, or revenue cycle domain experience (helpful but absolutely not required).