Senior Python Engineer

What you’ll do

Own Python services end-to-end, design, build, ship, and operate Lambdas and APIs (FastAPI) that the purchases investigation agents depend on 24/7.

• Reason about a fleet of services, not a single app, understand how dozens of Lambdas and services hand off work via API calls, SNS topics, SQS queues, and Kafka streams, and keep that mental map current as the system evolves.
• Master the messaging patterns, NS fan-out, SQS with DLQs and redrive, idempotency, retries, ordering guarantees, back-pressure, and the subtle differences between "event notification" and "command" flows.
• Debug across service boundaries, trace a single order or banking event from Kafka/SNS ingress through Lambda transforms, DynamoDB writes, Elasticsearch indexing, the data API, and into the webapp column an agent is looking at. You are comfortable reading CloudWatch Logs, Datadog traces, and correlating IDs across five hops to pinpoint which layer owns a bug.
• Translate between business and engineering, absorb requirements from risk, product, and CX stakeholders, turn them into concrete designs, and explain back to a non-technical audience what we shipped, why it matters, and how it helps them do their job.

What we're looking for

Must-have

• Strong Python (3.11+), production experience with async code, typed models (Pydantic).
• Hands-on with FastAPI or an equivalent Python web framework, including auth, validation, and operating it behind API Gateway.
• AWS serverless experience: Lambda, SNS, SQS (with DLQs), API Gateway, CloudWatch Logs and metrics, IAM, and DynamoDB
• Experience with Elasticsearch (query DSL, index lifecycle, tuning, operating clusters) at non-trivial scale.
• Strong stakeholder communication: you can sit with a non-engineer, understand what they need, and later explain what you built in plain language.

Nice-to-have

• TypeScript / React, you don't have to ship features in the webapp, but you're comfortable tracing bugs into the frontend and making small changes to skan-webapp.
• Java, some upstream Klarna services that we integrate with are written in Java; being able to read and reason about them speeds up cross-team debugging