About the company

Taste Labs is building the data and infrastructure layer for taste.

Our goal is to end AI slop. To make AI feel right, not just be correct. We raised $18.5M in seed co-led by Amplify and CRV, and most frontier labs are already customers.

AI has nailed objective domains and can generate anything. The hard part left is judgement: what fits, what feels like you, what's actually GREAT. We're turning that into something measurable, starting with design.

We do it on two sides: building the post-training data and RL environments that teach taste to frontier models, and the context and verification tools agents need to produce work that's more creative, more on-brand, more right.

If that problem excites you, you'll like it here!

What you'd own

You're building and running community as an operating system.

Supply chain operations for a curated expert network. Not community management. Not recruiting. The system that makes the network appreciate with use.

This is an execution seat. You run the machine, keep the data clean, and tell us where it's jamming. Scoped deliverables, owned end to end, on your schedule as long as the cadence holds.

You'll join community operations work that is already underway, and help build the structure as you run it.

Day to day

1. You keep supply data current. Who we have, what they're good at, where they sit in the tier model, whether they're available.

2. You publish the weekly funnel numbers: apps → portfolio yes → tested → passed → contracted → staffed.

3. You run onboarding mechanics. Schedule sessions, chase stalls, move experts through the steps, flag single-person bottlenecks before they cost a week.

4. You run reactivation campaigns. Send the sequences, track replies and bookings, route conversations to the right person, report what converted.

5. You run bridging. The between-project touchpoints that keep experts engaged and warm, on the cadence the Lead sets, so starts are fast and retention holds.

What you walk into

1,000+ vetted experts across design styles and specialties. A working but partial supply graph. A pod lead model in development. An ops team that has built real infrastructure and is building more. A clear thesis: community is a compounding asset, quality is institutionalized, and retention is structural rather than transactional.

You're not starting from zero. You're not inheriting something finished either. You're joining mid-build.

What we screen for

Doing vs drifting.

When you say you “owned” something, what were you allowed to decide without asking permission? Who didn't report to you that you had to align? What did you do when no playbook existed? We are hiring people who have built systems other people use, not people who have executed inside systems someone else built. The story we want is how you designed something that worked without you in the room.

Where your ego sits.

Ego in “my function” produces silos. Ego in “the mission” produces interfaces, documented handoffs, and data the next person can trust. We are hiring for the second.

Comfort in ambiguity.

The workflows you execute this month are being designed this month. If you need the playbook finished before you can move, this is not the seat.

The taste question.

We'll ask whether you have taste in something. Not whether your taste matches ours. Whether you have the muscle at all. Because the people you're emailing, scheduling, and tracking are people who have it, and they can tell when the person on the other end doesn't care.

Where you might be coming from

We are intentionally open. A handful of profiles match the shape of this work, none of which are obvious by title.

• Ops or production coordinators at design-first agencies, studios, or production companies

• Talent coordinators at agencies, casting operations, or fellowship and residency programs

• Marketplace or community ops at companies where supply quality mattered more than supply volume

• Program coordinators who ran cohorts, bootcamps, or apprenticeships end to end

• Ops generalists or EAs at fast companies who built the tracker everyone ended up using

• Freelance producers who have kept many-headed projects on schedule without authority

What disqualifies

• You think of community management as content programming: events, newsletters, swag

• You can't distinguish between “building a community” and “building a guild”

• Your background is volume recruiting: fill rate over fit

• You need someone to chase you for the Monday numbers

• You produce polished work but slowly. Polished alone is not the bar.

What you get

Scoped, steady contract work with a clear owner and a clear bar. A vendor-grade toolkit (Airtable, Notion, mail tools) and the budget to use it. A network of 1,000+ working designers who will know you by name. If the seat grows, you grow with it.

How the loop runs

Screen, then founding team member to screen for operating tempo and systems thinking, then a culture and commitment conversation. References at the end.