Software Engineer, GenAI

📋 Description

• • Abridge AI is seeking passionate and innovative GenAI Engineers, at all levels, to join our mission-driven team and contribute to the advancement of healthcare understanding through cutting-edge artificial intelligence.
• • In this pivotal role, you will be instrumental in translating the latest advancements in Large Language Models (LLMs) into robust, reliable, and real-world clinical systems.
• • You will collaborate closely with a dynamic, cross-functional team comprising AI researchers, practicing clinicians, and fellow engineers, fostering an environment of shared expertise and innovation.
• • The core of your work will involve designing and implementing sophisticated LLM-driven workflows that are capable of complex reasoning within intricate clinical contexts.
• • This includes developing and integrating agentic capabilities, enabling structured tool use, managing intricate branching chains of LLM calls, and ensuring seamless integration with advanced retrieval systems.
• • A key objective is to consistently generate outputs that meet the highest standards of clinical reliability, accuracy, and trustworthiness, directly impacting patient care and clinician efficiency.
• • You will play a critical role in developing and applying rigorous evaluation frameworks, encompassing both automated metrics and human-in-the-loop assessments.
• • These frameworks will be essential for continuously monitoring and improving the accuracy, robustness, and multilingual capabilities of our GenAI systems.
• • You will be tasked with designing experiments to meticulously probe potential failure modes, simulate challenging edge cases, and stress-test LLM workflows under realistic operational loads and demanding real-world conditions.
• • A disciplined, data-driven approach will be paramount in understanding and predicting model behavior.
• • This involves developing sophisticated tools to accurately measure system performance, conducting controlled A/B tests against established baselines, and generating clear, actionable insights to guide critical deployment decisions.
• • This high-impact role carries the responsibility for the end-to-end productionization of LLM workflows.
• • This includes deploying models into low-latency, high-uptime production environments, building comprehensive monitoring and observability systems, implementing robust post-processing guardrails, and managing efficient workflow versioning.
• • You will be responsible for designing and building advanced GenAI systems that transform LLMs into composable, dependable tools, effectively leveraging retrieval augmented generation, sophisticated tool use, agentic reasoning, and structured output formats.
• • Close collaboration with Machine Learning and Infrastructure engineers will be essential to scale and optimize these GenAI workflows, with a keen focus on managing critical parameters such as latency, context window utilization, and optimal model selection.
• • A significant aspect of the role involves writing high-quality, modular, and maintainable code that is resilient to failure, adaptable to evolving requirements, and conducive to rapid iteration and development.
• • You will be empowered to drive major architectural decisions, shaping how we architect our LLM workflows, define data flow pipelines, implement effective caching strategies for intermediate states, and structure the generative outputs for maximum utility and clarity.
• • The role demands a commitment to rigorous evaluation, including the creation of comprehensive benchmark datasets, the development of both automated and human-in-the-loop evaluation frameworks, the design of experiments to uncover failure modes and edge cases, the execution of A/B tests to inform deployment strategies, and the distillation of actionable insights from clinician feedback to guide continuous model improvement.
• • You will have the opportunity to leverage frontier capabilities by rapidly prototyping with the latest models, exploring new model capabilities, utilizing open-source tools, and experimenting with novel prompting techniques to push the boundaries of what's possible.
• • This position offers a unique chance to be at the forefront of generative AI in healthcare, making a tangible difference in how medical information is processed and understood, ultimately improving patient outcomes and clinician experiences.