AI Engineer

Apify is the largest marketplace of tools for AI. On Apify Store , tens of thousands of Actors , purpose-built tools for the web, help people and AI agents get real-time data, track competitors, generate leads, and connect apps. Anyone can build one and earn from it.

Join us to help people put the web to work. Apify can find missing children , protect consumers from fake discounts across the EU , and feed data to AI chatbots .

Working on our mission to build the largest marketplace of tools for AI, we're becoming one of the pioneers in MCP dynamic tool discovery and agentic payments via Coinbase's x402 standard , letting autonomous agents discover the right tools and pay for them on their own. Because we serve both humans and machines, we're also making natural language a first-class interface to the platform.

We stay close to where this is happening: we show our work at conferences like the AI Engineer World's Fair, where we spoke on the emerging agentic economy , and the MCP Dev Summit; we host our own Prague Crawl, where Jan Curn made the case that "MCP doesn't suck, your agent does" ; and we regularly rotate people through San Francisco. We're looking for an AI Engineer to help build all of this. It's a small team with a lot of surface area, so you'll own real problems end-to-end.

What you'll be working on:

Build and maintain the Apify MCP Server, turning tens of thousands of Actors into tools that agents can reliably call.

Keep the server ahead of a protocol that changes fast and often breaks: stateless transports, authorization hardening, and tool discovery.

Shape the agent behind Apify AI, from the runtime it runs on to the tools it reaches and the rough edges that surface when it runs a full task end-to-end.

Help autonomous agents discover and pay for Apify tools.

Build evals that turn real production traces into tests, catch regressions in CI, and tell us whether a change made the product better or worse.

Who we're looking for:

3+ years building real software, with backend experience in Node.js/TypeScript or Python.

You've built something on top of LLMs and know the pain firsthand: tools the model calls wrong, context that balloons, agents that confidently do the wrong thing.

Curious and self-directed: comfortable with loosely defined problems and keeping up with a field that changes every week.

You use the Apify platform, or want to, and care about building tools other people enjoy using.

Good communication skills in English.

Our tech stack:

Infra: AWS Compute (Kubernetes (EKS), EC2, Lambda), Helm, Terraform, CloudFormation, ArgoCD, MongoDB, Redis, DynamoDB, S3, GitHub Actions

Monitoring: Grafana, Prometheus, OpenTelemetry, Mezmo, Sentry, PagerDuty

Frontend: React.js, styled-components, Storybook, Chromatic, Cypress, Playwright

Backend: TypeScript/Node.js, Nest.js, Next.js, Docusaurus, Vitest

Tools: GitHub, Notion, Google Workspace, Slack

OS, code editor, and AI assistant of your choice (Claude Premium with unlimited tokens, GH Copilot, Cursor, Gemini, or JetBrains AI)

Process: two-week sprints, code reviews, tests, automating whatever we can, and deploying multiple times per day