Forward Deployed AI Engineer (Mirabel)

Forward Deployed AI Engineer

Enlightened Today is a rapidly expanding and innovative company leading the charge in the spiritual and wellness space. We are dedicated to changing the industry with bold, creative marketing strategies and forward‑thinking technical solutions. Our team thrives on collaboration, innovation, and crafting content that deeply resonates with our audience. Our business model centers on creating authentic connections with our audience through digital channels and delivering premium services and products that meet their needs.

The role

This is a forward‑deployed engineering role. Instead of sitting behind a backlog and building to spec, you’ll embed directly with the founders and the team, figure out what the business actually needs (which is often not what was first asked for), and ship working AI solutions fast. You’ll talk to the people who’ll use what you build, iterate in front of them, and turn rough ideas into things that work. In short: part engineer, part solutions architect, part builder‑who‑talks‑to‑humans. High autonomy, fast prototyping, strong communication, and a bias toward shipping something useful over perfecting something slowly.

This is a "hire for talent" role, and we care far more about what you can build and how you think than about your years of experience or pedigree. If you’re early‑career but exceptional, or senior and want to get your hands dirty again, both are welcome.

The kind of work

Your projects will change as the business does, but they’ll cluster around a few themes:

• Building and orchestrating AI agents to automate real work across engineering, operations, and marketing.
• Creating personalized, AI‑driven experiences for our customers: systems that understand context and respond in a genuinely helpful way.
• Turning data and signals into useful systems by taking what we know about a customer and making it drive better experiences and smarter marketing.
• Building internal tools that give a small team outsized leverage.

You’ll rarely get a finished spec. You’ll get a goal, some context, and the freedom to figure out the rest.

Our stack

• AI / LLM: Claude (API + Claude Code), multi‑agent orchestration, prompt and skill authoring, voice tooling
• Primary language: Python
• Backend / data: Postgres / Supabase (incl. pgvector), REST APIs
• Frontend: Next.js, React, Vercel, HTML/CSS
• Marketing infra: Klaviyo, Checkout Champ, Hyros, attribution and CRM tools
• Tooling: Git/GitHub, CI, modern AI‑assisted dev workflows

You don’t need every item on this list. You do need to be the kind of person who picks up a new tool quickly and ships with it.

What we’re looking for

Mindset

• Forward‑deployed instincts: you can sit with a non‑technical stakeholder, diagnose what they actually need, and turn it into something working, fast, and without a detailed spec.
• Strong communication: you’ll work directly with the founders and explain technical work to non‑engineers clearly.
• A bias toward shipping: working software over perfect software, and the judgment to know when each matters.
• Self‑direction: you thrive with autonomy and ambiguity rather than needing tightly defined tasks.

Core engineering

• Strong Python and solid full‑stack fundamentals, comfortable across backend logic, data, and a functional frontend.
• Solid engineering discipline: testing (we value TDD), clean Git workflows, and the instinct to build quality gates rather than skip them.
• Comfortable in shell and TypeScript/Node when the tooling calls for it, and you don’t have to be an expert, but you can’t be allergic to it.
• Data fundamentals: modeling data well, integrating with third‑party APIs, and knowing whether what you built actually moved the numbers.

AI / LLM

• Real, hands‑on LLM experience: prompting, retrieval (RAG), evaluating outputs, and handling the messy reality of non‑deterministic systems.
• Agentic experience: you’ve built or seriously operated multi‑agent systems and understand orchestration, tool/function calling, and review/eval loops.
• Fluency with AI‑assisted development: you use tools like Claude Code daily and know how to get great output from them.

Judgment

• Care around sensitive, emotional conversations. Our customers come to us about love and relationships; AI that interacts with them needs guardrails, scope limits, escalation paths, and clear boundaries.
• Care with personal data: customer information is sensitive (consent, data minimization, secure storage, CCPA).

Nice to have

• Prior forward‑deployed, solutions‑engineering, founder, or consulting experience, or anything that put you face‑to‑face with the people you build for.
• Conversational or voice AI experience (TTS/STT, low‑latency streaming).
• DTC e‑commerce, subscriptions, or performance‑marketing background.
• Familiarity with agentic frameworks (LangGraph, CrewAI, AutoGen, or similar) and the Agent Skills pattern.
• Recommendation or personalization experience.
• A portfolio of shipped products, repos, demos, or AI projects we can look at.

Why this role is different

• Genuine ownership. Direct access to the founders, no layers of approval, and the ability to ship to real customers quick.
• You’re close to the problem. You won’t be handed tickets in isolation. You’ll sit with the business, see the real need, and build for it directly.
• AI‑native by default. You’ll orchestrate AI agents rather than type every line, the way the best engineers will work in a few years.
• Breadth. If you’re tired of being a cog on one narrow feature, this is the opposite. You’ll touch the whole business.