Role Overview
Software Engineer 3 at MongoDB, extending Voyage’s embedding models to third‑party cloud platforms (AWS, Azure, GCP) and self‑managed environments, ensuring consistent behavior.
Responsibilities
- Port and tune the model server for Voyage embedding and reranking models.
- Productionize new models for delivery beyond Atlas, managing packaging and deployment on AWS, Azure, GCP.
- Design validation to prove third‑party deployments match first‑party behavior.
- Build operability with logging, metrics, health checks using Prometheus and OpenTelemetry.
- Debug issues across model servers, containers, configuration, and partner clouds.
Requirements
- 4+ years building backend, infrastructure, CI tooling.
- Strong skills in Python or Go, focus on performance and reliability.
- Experience with AWS, Azure, or GCP, containers, Kubernetes.
- Comfortable debugging across code, runtime, model servers, config, external deps.
- Familiarity with ML model serving and inference runtimes.
- Clear communication and cross‑team collaboration.
- Motivated to run Voyage AI models reliably for customers.
Nice to Have
- Experience deploying ML model servers (vLLM, Triton, TGI).
- Familiarity with Prometheus, Grafana, OpenTelemetry.
- Shipping software via cloud marketplaces or partner channels.
- Contributions to open‑source ML serving infrastructure.
Benefits
- Own how Voyage’s AI models reach customers outside first‑party Atlas.
- Work on model servers, inference, CI, observability, cloud integration.
- Collaborate with ML and platform teams to bring new models to market.
- Join a team valuing ownership, pragmatism, technical judgment.