The opportunity
As a Software Engineer on Collections Infra, you’ll help scale the infrastructure behind Notion’s database block. Databases power views, boards, charts, forms, and the workflows customers use to run their companies; as Notion grows into larger enterprise deployments and…
What you'll do
By your 90th day, you’ll understand how Collections Infra supports Notion’s: database block, ship improvements to database performance or reliability, and build context on the scale challenges behind larger enterprise workspaces and AI-driven write volume.
You’ll use AI as a real engineering collaborator: to learn unfamiliar systems faster, explore debugging paths, generate tests, summarize traces or logs, and pressure-test implementation plans before shipping.
You’ll contribute to the team’s work to make databases faster and more: reliable at P95 scale, especially as customers build larger centralized databases with more concurrent usage.
By the end of your first year, you’ll have helped raise the quality bar for: Collections infrastructure through better architecture, observability, documentation, or system boundaries that make future work easier to build on.
You’ll partner with senior engineers and product-facing teams to evolve: Collections as a platform, balancing internal developer needs with the external customer experience of databases.
Strong backend engineering fundamentals, with interest in performance,: reliability, scalability, and systems that need to hold up under real customer load.
What they're looking for
- Curiosity about distributed systems, microservices, and data-layer: trade-offs; experience in these areas is helpful, but we care most about your ability to learn quickly and reason from first principles.
- Practical comfort with JavaScript or TypeScript, or the ability to ramp: quickly in a codebase where those tools matter.
- Grit and truth-seeking in technical problem solving: you can iterate when the first approach does not move P50/P95 performance, and you’re motivated by finding the real bottleneck.
- A self-guided learning style: you use docs, code, teammates, and AI to build context, ask better questions, and turn ambiguity into steady progress.
