AI Engineer

This role sits at the intersection of AI engineering and application development. You’ll build customer-facing features from scratch, while also helping us integrate and scale the AI infrastructure that power them. You're deeply fluent in TypeScript and React, strong enough on the backend to own a feature end-to-end, and have meaningful experience building AI applications. You craft fast, polished, reliable software and you're drawn to problems that are genuinely unsolved.

We're looking for full stack engineers and those who lean more towards frontend.

Full stack. You understand applied AI well past the API layer: building systems that stay reliable when the model is unpredictable, writing evals that surface real signal, designing agentic workflows that hold up under production conditions. You've shipped RAG, tool calling, memory, or multi-agent patterns to real users and you know where each one breaks down.

Frontend. You notice things other engineers don't. The easing curve that's slightly off. The transition that happens too fast to register. The empty state that was clearly an afterthought. In a product where AI does most of the heavy lifting, the interface is where trust is built or broken — and you hold that line. You've built design systems from the ground up and have a point of view on motion, interaction, and what makes an AI-generated surface feel considered rather than random.

• Production-grade React, NextJS, and TypeScript
• Strong backend experience: async systems, API design, schema decisions, scaling under real load
• For full stack: applied AI in production like RAG, tool calling, memory, multi-agent patterns, and the ability to write evals and reason about whether a system is actually improving
• For frontend: design systems, animation and motion, data-fetching patterns (TanStack Query, tRPC), and deep care for browser fundamentals and performance
• Familiarity with our infrastructure and AI tooling (Cloudflare Workers/Durable Objects, Postgres + PGVector, LangSmith, Inngest) is a plus

Features that are fast, reliable, and polished. Architectural and interaction decisions that make the next thing faster to build. You raise the bar for the people working alongside you.

Your works speaks louder than your resume. We want to see how you think, what you've made, and how you work with AI. Our application asks for three things:

1. A Loom video (2-3 minutes)

Tell us who you are, what you're working on or have recently built, and how AI is a core part of how you work. Don't script it. We want to see how you think out loud.

2. Something you've built or designed 

A link, a side project, a campaign. Anything that shows what you're capable of making. One strong example beats ten weak ones.

3. A few short questions

1. Why Durable 
2. Where are you located
3. What are your salary expectations

We are a Vancouver-based company and candidates need to be based in Canada. Remote is fine - you can expect travel to Vancouver once per quarter (minimum) to work in-office with the team. 
Compensation is $180,000–$225,000 CAD base plus equity, for Vancouver-based candidates. Actual salary takes experience, location, and other factors into account.

Languages: Typescript, SQL, Bash, HTML5, CSS3, Python
Frameworks/Libraries: React, NextJS, React Query, React Native, Expo, TailwindCSS, RadixUI, DrizzleORM, trpc, NodeJS, Bun, WebRTC, PyTorch
Tooling: Turborepo, ESLint, Prettier, NPM/PNPM, Git, AI codegen
Databases: Postgres (including PGVector and JSON datatype), Object stores
Infrastructure: Docker, Cloudflare (Workers, Durable Objects, Pages, DNS, CDN), Vercel, Inngest, Render, Trigger.dev, Datadog, Together.ai, LangSmith, Hugging Face, Replicate, Langfuse
AI engineering: Large Language Models, Multi-Modal Models, Image Generation, In-painting, Voice Models
AI concepts (learn over time): Basic Prompting, Chain of Thought prompting, N-shot prompting, Prompting reasoning models, Tool calling, Preprocessing unstructured data, ReAct, Agent basics, Advanced agentic patterns, Evals, Memory, Generative UI, Streaming, Real-time, Multi-agent systems, Guardrails, Citations, Vector databases, RAG, Text embeddings, Knowledge graphs, Query routing, Synthetic data, Fine-tuning, RLHF, Diffusion Models, MCP, Computer use, Using/Serving Multi-modal OSS models