Job Overview:

You’ll work on our self-hosted LLM observability platform (built on Langfuse) — it’s where all our AI services send their traces, so we can see how every document and model call flows through the system, what it costs, and why things break. It’s a TypeScript/Node codebase running on Kubernetes (AKS). You’ll help build and operate the platform and the instrumentation around it — a great role for someone early-career who’s genuinely curious about how systems behave in production.

Job Responsibilities:

• Instrumenting our services with traces, metrics, and structured logs, and making that data useful in dashboards and alerts.
• Helping run and maintain the platform (web + worker services, and the queues and datastores behind them).
• Triaging issues by following a request end-to-end through the traces.

Requirements:

• 1–2 years of backend / platform / full-stack engineering or strong internship/project experience.
• Solid programming in TypeScript/JavaScript (Node.js) — or strong fundamentals in another language with genuine willingness to ramp on TypeScript quickly.
• Hands-on experience with at least one cloud provider (Azure, AWS, or GCP) and the basics of
• compute and networking — with a willingness to go deeper fast.
• Comfortable with Docker/containers (or eager to learn) and curious about Kubernetes.
• Genuine interest in observability — you like understanding how requests flow and why things fail, and you’re comfortable reading logs, metrics, and traces.
• Basic working knowledge of LLMs — what tokens, context windows, and quotas mean, and why tracing and cost-tracking matter for AI apps.
• Good debugging instincts; comfortable with Git, REST/HTTP, and JSON.
• Fluency with AI-assisted dev tools (Claude Code, Cursor, or similar), and a fast learner who likes to go deep.

Nice To Have:

• Kubernetes/Helm, or any container-orchestration exposure.
• OpenTelemetry, Langfuse, or distributed tracing experience.
• Next.js / React, tRPC, Prisma, or BullMQ.
• PostgreSQL, ClickHouse, or Redis.
• Azure specifically (AKS, ACR, Key Vault, Front Door), or IaC (Terraform/Bicep).
• CI/CD with GitHub Actions.