Senior Platform Engineer

TL;DR

We're rebuilding how companies hire — replacing gut feeling with science, on top of LLMs and agents. We're hiring a senior platform engineer in Stockholm to own the systems behind it: GCP infra, CI/CD, observability, data pipelines, and the AI tooling our engineers use daily. Broad role, real ownership, small team.

The role

Alva runs as event-driven microservices on GCP. We ship product features on top of LLMs and agents, and our engineers use Claude Code daily. The scope is broad — nobody we hire will have done all of it. We're hiring for judgment and how fast you pick up the rest.

What we're looking for

• You've operated production systems for several years and have opinions about how they should be built.

• You think in failure modes. You ask what happens at 3 am before you ask what happens on the happy path.

• You're effective with AI tools. You steer them, push back, and verify their output instead of trusting it.

• You explain decisions clearly and change your mind when someone has a better argument.

If you've done most of the work below in some form, at any cloud, at any scale, we want to talk. A checklist match isn't required.

What you'll do

Some shipped, some in flight, some still on the drawing board — a sample of what the team owns:

• Migrating our backend services from Kubernetes to Cloud Run, along with the Pub/Sub subscriptions.

• Introducing pgbouncer in front of Postgres to scale past our replica cap.

• Building a Backstage-style developer portal with an MCP layer — one source of truth for services, conventions, and architecture, for engineers and coding agents alike.

• Scaffolding agentic features so product squads ship them without reinventing prompts, evals, and guardrails each time.

• A protobuf schema registry and data contracts pipeline to fix the drift between services and BigQuery.

• Enabling the data team — shared patterns and pipelines (dlt, BigQuery) that let analytics work scale alongside engineering.

How we work

We use Claude Code heavily. We ship product features on top of LLMs (Gemini API), agents, and tooling like Langfuse for evaluation. We want engineers who treat AI as a collaborator they steer, not a tool they trust blindly.

The platform team's work touches every other team. What you ship is visible across the company in days.

Logistics

• Location: Stockholm, hybrid.

• Team: Small platform team reporting to the platform lead.

• Stack: Python, GCP (Cloud Run, GKE, Postgres, Firestore, Pub/Sub, BigQuery), Terraform, GraphQL/Apollo, dlt, Sentry, Gemini API, Langfuse.

Nice to have

None of these are required, but any one of them would let you carve out a specialty fast:

• Postgres depth: query plans, indexing strategy, connection pooling.

• Agent scaffolding and AI tooling: reusable patterns for prompts, evals, guardrails, MCP servers. Spec-driven approaches a plus.

• Distributed systems patterns: rate limiting, backpressure, queue throttling, handling bursts in a Pub/Sub-heavy system.

• Observability depth: distributed tracing, OpenTelemetry, performance profiling across services.

• Data engineering: dlt, dbt, BigQuery modeling. Not our daily focus, but useful background.

Probably not the right fit if

• You want a specialist role with one technology.

• You prefer long design cycles and heavy process over shipping and iterating.

• You'd rather not use AI tooling in your daily work.

Why Alva

At Alva Labs, we're rebuilding how companies hire — replacing gut feeling with science. We're building an AI-first hiring system that helps companies define roles, assess candidates, and make better decisions.

We're a small, ambitious team building from Stockholm. We value ownership, velocity, and low-ego collaboration.