Bellroy's search for radical effectiveness means we're always after better ways to get things done, and right now nothing is changing how work happens faster than AI. Tools that didn't exist a year ago are already part of how our crew designs, writes, analyses and builds. But there's a wide gap between having these tools and using them well. That gap is yours to close.
As our AI Enablement Lead, you'll be the steward of AI at Bellroy: the tools we choose, the practices we share, and the value we get from both. You'll run workshops that lift the whole crew's capability, sit with people while they untangle real problems with these tools (and tell them, kindly, when AI is the wrong tool for the job), and keep a clear-eyed view of what our AI toolset costs versus what it returns, recommending changes to tools and practices when the numbers or the experience say we should.
This isn't a role for someone who just loves shiny new things. It's for someone who can hold genuine enthusiasm for what these tools make possible alongside a healthy scepticism for hype and who gets real satisfaction from watching someone else have the "oh, that's how I should do it" moment.
You'll work right across the business: Creative, Marketing, Technology, Finance, Customer Support and more — and partner closely with our Technology, Continuous Improvement and People teams to make sure capability building sticks.
- Deep, hands-on fluency with modern AI tools, from chat assistants to agentic workflows, plus the habit of staying current as the landscape moves (weekly, lately).
- The knack for building a workflow in your head. You can sit with a team whose work you don't know, break a vague "could AI do this?" into steps, and get to something that actually runs without pretending to be the expert in their domain.
- Designed and run workshops or training that changed how people worked afterwards, for audiences as different as designers, accountants and engineers.
- Can handle the odd interruption "Quick question, how would I do this?" is the best part of your day, and you can diagnose fast, show a practical path, or say honestly that a spreadsheet, a script or a phone call would serve them better.
- Comfort with numbers. You can track usage and licence costs across a toolset, build a sensible picture of return, and make the case for change when the value isn't there.
- Can win the trust of sceptics and the restraint of over-enthusiasts in equal measure
- Sound judgement on responsible use, covering data privacy, security, intellectual property and quality, and the ability to write guidelines people actually read.
- An even keel. Something you recommended in March will be superseded in June, in public, and you'll need to update your advice without taking it personally.
Even if you don't tick every box, if the idea of walking into an unfamiliar team's problem and mapping a way through it sounds like a good week, we'd love to hear from you.
- Built automations or light integrations yourself, whether that's APIs, scripting or agent frameworks, enough to know what's easy, what's hard and what's brittle.
- Worked in capability building, enablement, change management or continuous improvement.
- Spent time in retail or product-based businesses, or alongside creative teams.
- Spent an hour with someone in Finance on a monthly reconciliation, asked a lot of questions about what happens between two of their steps, and built something that worked on the third attempt.
- Packaged a task the team keeps repeating into a reusable skill (with evals!) anyone can run and contribute to
- Build next month's training plan off the back of that, because the questions people ask in a 1:1 tell you more about what to teach than any survey will.
- Run a workshop for the Creative team where two people came back the next day wanting more, and one told you they still prefer their old way. Filed that away as useful.
- Talked a very keen colleague out of automating something that should stay human, and left them still keen.
- Sat with our Talent Director to move a workforce planning prototype out of a Claude Project and into a deployable Google Apps Script tool
- Pulled the licence report, found a batch of seats nobody had opened in the last month, and sat with our administrator to tidy them up.
- Rewritten a section of our internal guidelines because a model update quietly broke your own worked example.
- Answered a dozen Slack questions, three of which turned into proper projects.
Full-time, based at our Collingwood HQ. There's some flexibility for a hybrid arrangement (up to one day a week from home), but this role thrives on being in the room with the crew — so it's not set up as a remote position. Start date: whenever you're ready to jump in.
