Staff Backend Engineer (Ambient AI)

Athelas; United States; On-site; USD 140,000 - 210,000; Full time

Job summary

Athelas is building a next-generation ambient AI platform for healthcare, handling long-form audio, real-time streams, and AI model orchestration. The Staff Backend Engineer will design robust distributed services across ingestion, processing, inference workflows, and storage, with strong emphasis on reliability, observability, and secure data handling.

You will collaborate with product and clinical teams to deliver production-ready workflows, tuning performance, latency, and resilience for

Qualifications

• Distributed-systems concepts: idempotency, consistency, concurrency, backpressure, retries, failure isolation, eventual completion
• Experience operating services in a cloud environment: deployment, monitoring, scaling, incident response
• Ownership mindset and bias for action: drive infrastructure work through completion
• 8+ years of professional backend or infrastructure engineering experience
• Experience designing APIs, service boundaries, and data models for complex workflows
• Experience with asynchronous processing, message queues, event-driven systems, or durable workflow execution
• Strong debugging skills across application, infrastructure, data, and dependencies
• Ability to reason about real-time and long-running workloads with different latency/durability needs
• Track record of improving reliability, observability, scalability, or operational efficiency
• Experience with containers, Kubernetes, IaC, and cloud-native deployments
• Familiarity with observability tools: Grafana, OpenTelemetry, Sentry
• Experience defining SLOs, error budgets, alerts, and production incident response
• Experience building internal tooling for workflow inspection, replay, evaluation, or model debugging
• Experience with model routing, inference gateways, rate-limiting, batching, caching, or GPU workloads
• Experience with event-streaming/messaging systems: Kafka, Pub/Sub, SQS
• Familiarity with workflow orchestration systems like Temporal
• Experience orchestrating large language models or multi-model inference pipelines
• Experience building speech-to-text, diarization, transcription, or audio ML workflows

Responsibilities

• Design, build, and operate backend systems powering Ambient AI products across mobile and web
• Develop and evolve inference workflows coordinating transcription, diarization, LMs, clinical extraction, summarization
• Build orchestration for long-running, multi-stage workflows with scheduling, retries, timeouts, and dead-letter handling
• Ensure idempotency, replay, and safe workflow recovery
• Define model routing, versioning, and configuration; track progress and workflow state
• Establish strong observability with logging, metrics, tracing, dashboards, and alerts
• Lead incident response and post-incident analysis to drive improvements
• Collaborate with mobile, product, AI, security, and clinical teams to productionize capabilities
• Improve cloud infra, deployment, capacity planning, and tooling as platform scales

Job description

• We’re building a next-generation ambient AI product suite for healthcare, helping clinicians focus on patient care by automatically capturing, understanding, and structuring clinical conversations
• Behind that experience is a distributed system responsible for ingesting long-form audio, processing real-time and asynchronous streams, orchestrating multiple AI models, and reliably producing clinical outputs
• These workflows must remain observable and recoverable across network interruptions, partial failures, model timeouts, and rapidly changing inference infrastructure
• We’re hiring a Staff Backend Engineer to help build and operate this foundation
• You’ll work across audio ingestion, media processing, inference orchestration, distributed workflows, storage, observability, and cloud infrastructure
• This role is ideal for an engineer who enjoys making complex systems dependable
• You think carefully about failure modes, idempotency, backpressure, data integrity, latency, and operational simplicity
• You also understand that infrastructure exists to serve the product: the systems you build must ultimately create a fast, reliable, and trustworthy experience for clinicians
• Design, build, and operate backend systems that power our Ambient AI products across mobile and web
• Build and evolve inference workflows that coordinate transcription, diarization, language models, clinical extraction, summarization, and other AI capabilities
• Develop orchestration systems for long-running, multi-stage workflows, including:
• Scheduling, queueing, and workload prioritization
• Retries, timeouts, fallbacks, and dead-letter handling
• Idempotency, replay, and safe workflow recovery
• Model routing, versioning, and configuration
• Progress tracking and user-visible workflow state
• Graceful degradation when dependencies are unavailable
• Improve the reliability and scalability of services operating under variable, compute-intensive workloads
• Define and maintain service-level objectives for critical workflows, including availability, processing latency, completion rates, and data durability
• Build strong observability across services and pipelines through structured logging, metrics, distributed tracing, dashboards, alerting, and diagnostic tooling
• Lead incident response and post-incident analysis for important production failures, turning operational lessons into lasting improvements
• Identify systemic failure patterns and eliminate them through better abstractions, automation, testing, and architecture
• Improve cloud infrastructure, deployment systems, capacity planning, and operational tooling as the platform scales
• Design APIs and data models that allow mobile, web, and internal systems to interact with long-running workflows safely and predictably
• Build tools that make inference pipelines easier to inspect, test, replay, evaluate, and debug
• Balance reliability, latency, quality, and infrastructure cost across AI and media-processing workloads
• Partner closely with mobile, product, AI, security, and clinical teams to turn new capabilities into production-ready workflows
• Mentor other engineers and help establish strong practices for distributed systems, observability, operational readiness, and backend architecture

• A strong understanding of distributed-systems concepts such as idempotency, consistency, concurrency, backpressure, retries, failure isolation, and eventual completion
• Experience operating services in a cloud environment, including deployment, monitoring, scaling, and incident response
• An ownership mindset and bias for action. You identify important risks, create clarity, and drive ambiguous infrastructure work through completion
• Experience designing, building, and operating production distributed systems8+ years of professional backend or infrastructure engineering experience
• Experience designing APIs, service boundaries, and data models for complex product workflows
• Experience with asynchronous processing, message queues, event-driven systems, or durable workflow execution
• Strong debugging skills across application, infrastructure, data, and external dependency boundaries
• An ability to reason about both real-time and long-running workloads with different latency and durability requirements
• A track record of improving system reliability, observability, scalability, or operational efficiency
• Comfort working across backend, infrastructure, product, and AI systems rather than within a narrowly defined layer
• Strong proficiency in at least one modern backend programming language
• Exposure to healthcare, HIPAA, SOC 2, encryption, privacy, security, or other regulated environments
• Experience supporting AI-native, real-time, or agentic product experiences
• Experience with offline clients, background synchronization, or systems that reconcile delayed and duplicated events
• Experience optimizing systems for throughput, tail latency, infrastructure cost, and workload isolation
• Experience with containers, Kubernetes, infrastructure as code, and cloud-native deployment systems
• Familiarity with distributed tracing and observability platforms such as Grafana, OpenTelemetry, Sentry, or similar tools
• Experience defining SLOs, managing error budgets, designing alerts, and leading production incident response
• Experience building internal tooling for workflow inspection, replay, evaluation, or model debugging
• Experience with model routing, inference gateways, rate limiting, batching, caching, or GPU-backed workloads
• Experience with event-streaming and messaging technologies such as Kafka, Pub/Sub, SQS, or similar systems
• Familiarity with workflow orchestration systems such as Temporal or similar durable execution platforms
• Experience orchestrating large language models or multi-model inference pipelines in production
• Experience building speech-to-text, diarization, transcription, or other audio-based machine-learning workflows