Senior Software Engineer, Headway

• A senior engineer who combines deep cloud and infrastructure expertise with the instincts to drive alignment across teams
• You thrive in ambiguity, operate with a high degree of ownership, and are as comfortable setting technical direction as you are building
• You make other engineers better through architecture reviews, runbooks, and paved-road tooling that raises the baseline for everyone who ships on the platform
• 5 or more years in platform, infrastructure, or SRE roles at companies running significant production traffic
• Deep AWS expertise and production ownership of compute and networking at scale (ECS, EKS, RDS, networking, IAM)
• Strong infrastructure-as-code experience, particularly Terraform, including designing self-serve platforms for other engineering teams
• Hands-on autoscaling and capacity engineering, and container orchestration with ECS and/or EKS
• Track record making deploys safe and self-serve for other teams, not just your own

Desirable

• Building a new mental healthcare system at Headway is only possible because of the scale and leverage that software can provide
• Infrastructure Engineering is one of three teams in Engineering Foundations, alongside Agentic Engineering and Eddy, Headway's internal AI platform
• We own the infrastructure Headway runs on, as well as the performance and scalability of Headway's FastAPI/Python monolith, and the AWS infrastructure it runs on
• We are the paved road the rest of engineering builds on: databases, deploys, asynchronous platform, observability, autoscaling, and self-serve compute infrastructure
• As AI dramatically increases code velocity across the org, our mandate is to make every change safe from the start and build systems that get safer as they get faster
• You will be on the Infrastructure team, which owns the cloud platform Headway runs on
• You will be the senior engineer driving work to isolate blast radius, make autoscaling trustworthy, and build a self-serve infrastructure platform so that the Infrastructure Engineering team focuses on the strategic and unusual, not the routine
• Deployment architecture and blast-radius containment. Redesign deployment so that a mistake in one part of our service cannot block or take down others. Continue to drive our shift toward per-service deploy isolation and functional-area slices that contain failures rather than propagating them across the platform
• Container footprint and networking. Own the ECS and EKS footprint, evaluate broader EKS adoption for AI workloads, and design the next iteration of inter-service network connectivity
• Capacity and scaling. Own capacity for a spiky workload: floors computed ahead of demand rather than chased by reactive scaling, self-deriving from data, with drift caught early
• Self-serve infrastructure platform. Build the Terraform self-serve platform with guardrails so engineering teams own their standard infrastructure changes and Reliability Engineering reviews only the non-standard ones
• Cloud cost attribution and controls. Stand up per-team cost attribution across AWS, Datadog, and LLM spend. Make infrastructure costs visible and attributable so teams can make informed tradeoffs
• Python runtime and dependency health. Own how the monolith behaves under load: garbage collection, event loop contention, and the runtime limits that bite first. Lead the framework and package upgrades most teams defer