4 months ago
Base Salary
$200k - $280k/yr
Responsibilities
- Operate and evolve Vapi’s cell-based, multi-region infrastructure for 99.99% call completion.
- Manage regional EKS clusters spanning backend, networking, persistence, monitoring, models, and Kafka workloads.
- Operate and scale stateful systems including Postgres, Redis, Kafka, Temporal, etcd, and ClickHouse.
- Build Go control-plane services such as cluster-manager, traffic-control-plane, and environment-manager.
- Own infrastructure domains such as multi-region MSK, Postgres/Neon consolidation, or programmatic cluster creation with Cluster API.
- Lead roadmap initiatives including new-region launches, stateful workload migrations, and resolution of the SIP gateway single point of failure.
- Establish infrastructure standards, runbooks, failure-domain models, and capacity targets while supporting design reviews and incident response.
Requirements
- Production experience running multi-cluster Kubernetes on EKS, including backend, networking, persistence, monitoring, models, and Kafka clusters.
- Experience operating stateful data planes at scale, including sharding and data migration across Postgres, Redis, Kafka, Temporal, etcd, or ClickHouse.
- Fluency with Envoy and Cilium/eBPF, including debugging response flags, conntrack drops, and cross-zone load-balancer behavior.
- Production experience running multi-region Kafka on MSK, including regional topic naming, Pulumi drift, and compliance constraints.
- Ability to write Go for control-plane services and own code in a Go-based infrastructure stack.
- Experience with Cluster API or a similar system for programmatic cluster creation.
- SIP/RTP or telephony experience is a bonus.
- Cell-based or shard architecture experience is a bonus.
Benefits
- Competitive salary and equity ownership are offered, with no specific compensation figures stated.
- Medical, dental, and vision coverage.
- Quarterly company off-sites and flexible time off.
- Catered meals, transportation, gym access, and a $10,000 annual learning and development budget.
Tech Stack
