AI Engineer

Job Description

Reasonable Magic Engineer No. 1

What's Reasonable Magic?

Reasonable Magic is an AI product development studio. We bring together business strategy, design thinking, and AI engineering to build intelligent software systems with assessable business value at a speed previously unknown to mankind.

We see ourselves as builders in a proud lineage of ambitious, thoughtfully optimistic engineers, scientists, designers, and business people who prioritize real value creation over hype and bullshit. We work with clients who want AI products in production — not slide decks, not demos, not PoCs that die on the vine.

Constant evolution and sharp intuition aren't aspirational for us; they're prerequisites. We use AI in our own workflows and in the solutions we build for our partners — not so we can pay less attention, but so we can pay more of it to what actually matters.

What Are We Looking For?

An independent, curious, proactive problem solver who will collaborate directly with senior leadership to develop and deliver AI solutions. Someone with a strong opinion on the best AI-driven tools on the market and using them to execute real-world solutions with AI at the center. Someone to amplify our development capacity.

This isn't a "junior engineer takes tickets" role. We have Agents that do that. We are looking for someone who is capable of systems thinking, owns execution, and brings an informed opinion. The kind of person that checks their own work. You'll own meaningful pieces of development work and solution delivery from end-to-end, UX to deployment. You don’t have to know all the pieces of this to start, but one must be willing to learn and that is the long term expectation.

What Do We Need?

Must Have

1. Core Values Fit — see below. Non-negotiable.
2. Communication Discipline — you write things down, ask clarifying questions early, and raise risks before they become emergencies.
3. Software Engineering Fundamentals — you're a JS or Python native and literate in the other. You understand service and codebase design fundamentals and are passionate about developing your craft.
4. AI-Native Development Workflow — you use Claude Code, Cursor, Codex, or equivalent every day. You know how to drive an LLM to produce production-quality code, not just act as a typing assistant.
5. Cloud Fundamentals — Some past use of cloud services and deployments.
6. Frontend Taste — you don't have to be a designer, but you need some literacy of frontend concepts, taste in UI, and the AI savvy to pull those together into a decent page.

Nice to Have

1. Agentic Development Experience — you've shipped applications that depend on LLMs for core application logic. You understand tools, MCP, and context engineering, and you're familiar with LLM APIs and frameworks like LangChain, LangGraph, or custom orchestration with the Anthropic SDK.
2. Cloud Experience — GCP preferred: Cloud Run, Firestore/Postgres, Pub/Sub, IAM — the working set. AWS or Azure experience is transferable if you've actually deployed things.
3. Infrastructure-as-Code — comfort with Terraform or equivalent.
4. Design Sensibility — you can make a UI look fantastic, not just functional.

Reasonable Magic Core Values

1. Curiosity — the best engineers ask "why" three times. We assume the documented answer is incomplete and the obvious solution is wrong until proven otherwise. If you're not curious about how something works, you'll write code that surprises you in production.
2. Experimentation — we ship to learn, not to be right. Small bets, fast feedback, and a willingness to throw away yesterday's solution when today's evidence says otherwise. The AI stack rewards people who run the experiment instead of arguing about it.
3. Ambition / Growth Orientation — we want people who are getting better every quarter, not just executing on what they already know. The AI tooling landscape moves weekly; standing still means falling behind.
4. Optimism — building new things is hard. The default state of any project is "broken." If you can't hold the belief that the problem is solvable while staring at a wall of red errors, this work will grind you down.
5. Thoughtfulness — we'd rather think for ten minutes and code for an hour than the reverse. Premature implementation is the most expensive bug.
6. Radical Candor — in engineering, it worked or it didn’t , and we must always fight the urge to pretend. Care personally, challenge directly. Say the thing in the room, not in DMs after. We'd rather hear "this approach is wrong because…" than watch you nod along and ship something neither of us believes in. Politeness that hides the truth costs us all time.

How We Work

• Remote-first, async-default. Real-time when it matters.
• Contract-based to start, with a path to longer-term engagement if it's working for both sides.
• You work directly with me on real client problems, not isolated R&D side quests.
• Stack: GCP, Python, FastAPI, React/Vite, Claude/Anthropic SDK, Firestore + Postgres, Cloud Run, GitHub.

Application Process Overview

Our comprehensive selection process ensures we find the right fit for both you and our clients:

1. Initial Application - Submit your application and complete our prequalifying questions
2. Video Introduction - Record an video introduction to showcase your communication skills and work experience
3. Role-Specific Assessment - Complete a homework assignment tailored to the position (if applicable)
4. Recruitment Interview - Initial screening with our talent team
5. Executive Interview - Meet with senior leadership to discuss role alignment
6. Client Interview - Final interview with the client team you'd be supporting
7. Background & Reference Check - Professional reference verification
8. Job Offer - Successful candidates receive a formal offer to join the team

Each stage is designed to evaluate your fit for the role while giving you insights into our company culture and expectations. We'll keep you informed throughout the process and provide feedback at each step.