The role
Data labs built on site with the companies we work with, from the first conversation to production.
Each forward deployed engineer owns a deployment end to end: the first workshop, the architecture, the rollout on the company’s own machines, and every failure found in the field.
The stack runs from the Ambient desktop app to end-to-end encrypted inference on our own bare-metal GPU clusters, and to fully on-premise deployments for the most sensitive customers.
The work
- Scope deployments with customer teams and turn them into running systems.
- Ship across the stack: TypeScript, Svelte, Electron, Rust and model-backed services.
- Run confidential VLM inference in production: serving, enclave attestation, throughput.
- Own infrastructure on AWS and Terraform-based on-premise installs.
- Bring what breaks in the field back into the product.
The bar
- Years of full-stack TypeScript and projects owned from architecture to production.
- Production infrastructure on AWS or similar, with deep Terraform.
- Agent-native: swarms of coding agents are the daily tools.
- Calm and precise in front of demanding enterprise engineers.
- Nice to have: Rust, local-first systems, multimodal models, evaluations.