Founding Backend Engineer - LLM Orchestration

Healthcare documentation is broken. Nestmed is fixing it with an AI platform that gives clinicians their time back.

In just one year, we’ve scaled to support tens of thousands of clinicians across more than a million patient visits. We're now the trusted partner for over 60 home health agencies, including 7 of the top 10 enterprises in the US.

Our founding team—hailing from Stanford, YC, Google, and Meta—is backed by the founders of PayPal and Plaid to build the essential infrastructure for the future of the $500B home healthcare industry.

As the founding Backend Engineer on our LLM Orchestration team, you'll be deploying and managing LLMs at scale, learning how to orchestrate them in complex production scenarios that directly impact patient care. You'll rebuild and maintain our core AI inference engine that powers all of Nestmed's intelligent capabilities across several thousand clinical conversations daily.

Our system orchestrates over a dozen different AI models - both fine-tuned in-house models and third-party APIs - with low latency and high availability. You'll work on complex technical challenges like intelligent model routing based on clinical context, implementing sophisticated fallback strategies across multiple providers, optimizing inference costs through batching and caching, and ensuring clinical accuracy through comprehensive model evaluation pipelines.

This isn't about calling OpenAI APIs. You'll build sophisticated orchestration logic that selects optimal models for each clinical task, implements custom retry and circuit breaker patterns for provider failures, manages rate limits across multiple concurrent workflows, and maintains detailed performance metrics across the entire AI pipeline. You'll start as the solo engineer on this critical infrastructure and grow it into a robust team handling core AI engineering.

• Build and optimize our core AI inference engine that routes requests across multiple LLM providers based on clinical context, cost optimization, and latency requirements

• Design robust model serving infrastructure with intelligent load balancing, failover mechanisms, and A/B testing frameworks for model evaluation in production

• Implement production-grade AI pipelines with comprehensive observability, distributed tracing, and real-time performance monitoring for healthcare-critical workloads

• Optimize inference costs and latency through intelligent request batching, response caching, model quantization, and dynamic provider selection algorithms

• Build custom model fine-tuning and deployment pipelines for healthcare-specific tasks using frameworks like Transformers, vLLM, and distributed training infrastructure

• Create sophisticated prompt engineering systems that dynamically optimize prompts based on clinical context and historical model performance data

• Design comprehensive evaluation frameworks that continuously monitor model accuracy, clinical safety, and regulatory compliance across all deployed models

• Build model versioning and deployment systems that support safe rollouts, instant rollbacks, and controlled experimentation in production healthcare environments

• 6+ years of backend engineering experience building high-performance distributed systems, with focus on latency-critical applications and reliability engineering

• Deep production experience with LLMs including multi-provider orchestration, custom model serving, and building reliable inference infrastructure at scale

• Strong expertise in ML infrastructure including model serving frameworks (TensorRT, vLLM, TorchServe), distributed training, and GPU optimization

• Experience with model evaluation and monitoring including A/B testing frameworks, performance monitoring, and building comprehensive observability for ML systems

• Proficiency in Python and ML frameworks with hands-on experience in model fine-tuning, prompt engineering, and deploying custom models to production

• Track record scaling ML systems with experience optimizing inference costs, managing multiple model providers, and building reliable AI infrastructure

• Understanding of healthcare or regulated industries where model accuracy, auditability, and compliance are mission-critical requirements

• San Francisco-based and excited about working closely with AI researchers to productionize cutting-edge models for healthcare applications

You'll be building the AI infrastructure that processes millions of patient interactions, directly impacting care quality for thousands of patients daily. Every optimization you make reduces healthcare costs, improves clinical accuracy, and enables new AI capabilities that transform patient outcomes.

You'll start as the founding ML infrastructure engineer and build this into a world-class AI platform team. Join us in San Francisco to build the most sophisticated LLM orchestration system in healthcare alongside leading AI researchers and clinical experts.

If you’re passionate about building high-impact products that solve real-world problems, we’d love to hear from you. Apply today!