Software Engineer - Early Career

About the role

Abridge is seeking an early-career Software Engineer to join a mission-driven team building AI systems for healthcare. You will own the implementation of agentic LLM workflows that power real-time clinical documentation and decision support. This role centers on translating cutting-edge language model research into reliable, production-grade features used directly by clinicians at the point of care. You will be responsible for end-to-end feature delivery, collaborating closely with researchers, clinicians, and product teams to ensure solutions meet strict safety and usability standards. The position requires comfort with ambiguity and a bias for action as you help define processes for agentic coding and verification in a regulated environment. You will have a direct line of sight from your code to patient outcomes and clinical workflows. This is a hands-on role where your contributions will be visible, impactful, and foundational to the future of the platform.

Key facts

What you'll do

• Build and iterate on agentic LLM systems, including retrieval pipelines, structured tool use, and chained LLM workflows, working alongside a multidisciplinary engineering team.
• Contribute to evaluation frameworks that measure accuracy, robustness, and clinical reliability, helping maintain automated pipelines and human-in-the-loop review processes.
• Own pieces of both backend and frontend systems across Abridge, ensuring end-to-end functionality and performance in production environments.
• Think with an agent-first approach, continuously learning and relearning agentic coding patterns as tools and best practices evolve.
• Apply sharp judgment to determine where the human in the loop is critical, balancing automation with safety and oversight.
• Help prototype with new models, prompting techniques, and open-source orchestration tools such as LangChain and LlamaIndex to validate approaches quickly.
• Contribute to monitoring and observability systems that keep LLM workflows healthy in production, detecting issues before they affect clinicians.
• Collaborate across ML, infrastructure, product, and clinical teams, developing a deep understanding of the users and clinicians we serve.
• Learn from senior engineers on a daily basis, bringing curiosity and fresh perspective to solve complex technical problems under real-world constraints.
• Write and review code that powers core features, ensuring reliability, maintainability, and alignment with healthcare compliance expectations.
• Participate in design discussions, offering input on system architecture and user experience for features that impact clinical workflows.
• Ship production-grade features on a regular cadence, taking ownership of testing, rollout, and iterative improvement alongside cross-functional partners.
• Experiment with new tooling and methodologies to improve development velocity without compromising safety or correctness.
• Document technical decisions and system behavior to support long-term maintainability and knowledge sharing.

Requirements

• Hold a degree in Computer Science or a related field, or possess equivalent experience through projects, bootcamps, or open source contributions, with a focus on what you can build.
• Have hands-on GenAI experience through coursework, personal projects, internships, or hackathons, demonstrating practical exposure to LLM APIs, RAG pipelines, or agentic systems.
• Show familiarity with LLM orchestration concepts such as prompt chaining, tool use, and retrieval, even if you have not yet used them in a production setting.
• Use AI tooling as part of your daily workflow to write, debug, and learn, while knowing when to trust automated suggestions and when to verify outputs rigorously.
• Demonstrate curiosity about model evaluation, failure modes, and the practical steps required to make an LLM system reliable in clinical contexts.
• Work effectively in a collaborative, low-ego environment where fast execution and team support are essential.
• Thrive in a fast-paced startup environment where priorities can shift and ownership is expected at every level.
• Commit to working onsite five days per week in either the San Francisco or New York office, as stated in the role details.

Practical notes

• This role requires onsite work 5 days a week in either the San Francisco or New York office.
• Candidates must be eligible to work in the country where the position is based without requiring sponsorship at this time.
• No visa sponsorship is available for this role.
• The listed location is the San Francisco office, and candidates should be prepared to work from that site.
• No specific deadlines for application submission are communicated in the source material.