About the role
About Avra
Avra is a frontier AI lab building relational foundation models from the ground up (no third-party wrappers).
We model how businesses and individuals are connected and interact over time. Put simply: we bring connections into the math.
Leading enterprises in LatAm use Avra's models to power key decisions in credit, growth, and fraud. Our customers include the largest banks and marketplaces in Latin America, such as Mercado Libre, Telefonica, Banco Inter and Getnet.
Avra was founded in 2024 by Viviane Meister and Bruno Alano (ex-OpenAI). Our investors include NXTP, Clocktower Ventures, as well as the Sequoia and a16z scouts. We are accelerated by NVIDIA, AWS, and Google.
We are a team of builders. Join us.
About the role
At Avra, every technical IC is a Member of Technical Staff (MTS). The title doesn't put anyone in a silo: you own systems and outcomes, not steps in a function, and you keep building depth in your area.
In this role, you'll join the Platform team as our go-to expert on observability and reliability. Our customers make real-time decisions based on our responses, so when we're down, their operations stop. Avra's cloud is just one more dataplane, alongside the dataplanes we operate inside customer environments — so observability and reliability have to work the same way everywhere.
What you'll do
- Evolve our observability stack for logs, metrics, traces, and alerting.
- Make sure every dataplane, in our cloud and on-premise, reports its active release, health, heartbeat, logs, metrics, and usage to the control plane.
- Bring telemetry into customer clusters within a model where agents only make outbound connections.
- Detect drift between the desired state and what's actually running in each environment.
- Monitor the health of our deployment and runtime agents.
- Provide visibility into ephemeral workloads, such as the Ray clusters that run our batch inference.
- Define SLOs, lead incident response and postmortems, and reduce MTTR — including when a fix requires coordinating with the customer.
- Reduce telemetry cost: less redundant data, more useful signal.
How we measure success
- 99.9% serving availability, with incidents trending down.
- MTTR, including on-premise incidents.
- Near-zero drift between desired and actual state.
- All agents active and reporting, across every dataplane.
What we're looking for
- Deep experience with OpenTelemetry and observability backends.
- Hands-on practice with SLOs, error budgets, actionable alerting, and incident management.
- Strong experience with Kubernetes and infrastructure as code (Terraform / Helm ).
- Experience operating software in environments you don't fully control.
- Production-quality code and reviews, and a willingness to operate what you build.
Nice to have
- Shipping software to customer-hosted Kubernetes (e.g., Helm, outbound-only connectivity).
- GCP or GKE, AWS or EKS.
- ML multi-node/multi-cluster workloads in production.
- Financial services or regulated environments.