AI Enablement Specialist

Penguin Ai

About The Position

AI operations here means hands-on AI support and enablement: helping everyone at the company do better work with AI. Most days a teammate brings a problem that’s tangled, half explained, and urgent. The job is to get on a call, figure out what’s really going on, and leave them with something simple that works. Then it becomes a skill, a doc, or a tool so nobody has to ask twice. This is a support and enablement role at its core - if you enjoy helping people and you’re highly capable with tools like Claude Code, keep reading. If you mainly want to build product with minimal interaction, this probably isn’t the right fit. We define the problem before we automate it: an agent can’t solve something no human has figured out and clearly described first. That takes a clear view of what AI can’t do, real taste for building the smallest thing that solves the problem, the discipline to verify model output before passing it along, and the discretion to flag anything that could become a compliance or security issue - we work in healthcare.

Requirements

Clear, professional communication: You’re a fluent communicator, written and live, with engineers, executives, and people who’ve never opened a terminal - these matters more than anything else on this list
Live problem-solving: You’ve done remote tech support, consulting, or something like it, and a messy problem on a screen share doesn’t rattle you
Real AI fluency: You already use Claude Code, Claude Desktop, skills, and MCP for real work or for fun, and you’ve connected MCP servers and fixed them when they broke
Prompt and context engineering: You understand why an output is bad and how to fix the inputs
Microsoft 365 and Google Workspace: You’re fluent in both as a power user and light admin
Git and the command line: You use git daily and are comfortable working from the terminal
Fast, independent learning: You teach yourself and pick up new tools from docs and experimentation, often on your own time because you want to

Nice To Haves

You've not just survived but thrived in the beautiful chaos of a high-growth startup
Python or JavaScript/TypeScript, or full stack development exposure
Docker, and AWS, Azure, or GCP
HubSpot
Open source agent harnesses, or you have built your own harness or MCP server
Healthcare experience, or exposure to HIPAA, SOC 2, or similar compliance work
Formal training in software development or IT. Helpful, never required.

Responsibilities

Get on calls with teammates across sales, ops, engineering, and leadership. Learn their workflow, show them a better way, make it stick
Debug Claude Code, Claude Desktop, skills, and MCP connections in real time, often over screen share. Stay calm, think out loud, get them unblocked
Turn recurring requests into reusable skills, prompts, and workflows that chain several MCPs and services together (HubSpot, Microsoft 365, Google Workspace, Slack, Notion, and more)
Keep our shared context, prompts, and templates accurate, organized, and worth using
When a request is really a process problem dressed up as an AI problem, recognize it and bring it to Nick
Write the short guide, record the two-minute walkthrough, run the occasional session
Try new models, harnesses, and tools as they land, then talk through their limitations and potential opportunities