Build Systems ThatOutlive the Sprint

Design and ship retrieval and agent systems that hold up under real traffic, with evaluation harnesses, tool contracts and inference paths measured in milliseconds.

• Python
• LLM tooling
• Vector search
• Evaluation

Own the delivery path end to end: infrastructure as code, progressive rollouts, and the observability that tells us what broke before a customer does.

• Kubernetes
• Terraform
• Go
• Observability

Build event-driven services and durable workflows where correctness under load matters more than throughput on a benchmark.

• Go or Rust
• Kafka
• PostgreSQL
• Distributed systems

Take features from a rough brief to production: considered interfaces on top of backends that hold, shipped on a weekly cadence.

• TypeScript
• React
• Next.js
• Design sense

Cost-aware autoscaling, hardened pipelines and migrations that finish. You will inherit systems mid-flight and leave them better documented.

• AWS or GCP
• CI/CD
• Vault
• Networking

A paid rotation across AI, backend and platform work, paired with a senior engineer from week one. We hire from this programme.

• Curiosity
• Fundamentals
• Willingness to ask