Luna · The AI workplace
Welcome to the AI workplace.
Been here before? · skip the onboarding and go straight to your daily view
Go to DashboardThe five-step flow
The first-run experience.
Linear, on purpose. Each step ends in a real action you take — a chat, a click, a decision. Click through; it should feel like getting set up, not reading a wiki.
- 01
Welcome
You've been added to Luna's AI workplace. This is where every Luna employee works alongside the AI fleet — chat with shared agents, watch what they're doing, ask for one of your own. About 20 minutes total. You can stop and come back any time.
- 02
Meet Basal — your first agent
Basal is the shared assistant every Luna employee can use. Day-to-day questions about the company, drafting, summarizing, brainstorming, code snippets. Open a chat, ask it something real, and notice the response. Conversations are private to you; nothing is used to train any model.
- 03
See what else exists
There are six agents on the Luna AI Workforce today. Some are for everyone (Basal, Pen). Some are owner agents — built for one person (Clara). Some are platform-internal (Pump, Sensor — used by the team building this). Lancel is for clinical-data work and is gated to phi-authorized people. Browse the catalog and notice which ones are available to you.
- 04
Pick your role-specific deep-dive
Five role tracks below, ~20 minutes each. They cover what we're asking of you, what we're not, where AI fits in your loop, and what stays human. The first module on every track is the same: what we're not asking you to do — that's the boundary, written down. Pick the one that matches your work.
- 05
Get an agent of your own (optional)
If you find yourself doing the same multi-step thing twice a week — triage, drafts, lookups — that's a workflow worth giving to an agent. Pen runs the intake. Five minutes to file, a few days from request to working agent. There's no quota; everyone at Luna can ask for one.
Step 4 · Role tracks
Pick the track for your work.
Five role-specific tracks. ~20 minutes each. The engineering and clinical tracks share material; the ops and leadership tracks address how to use AI without setting wrong incentives for the people around you. If your work crosses two, do both — they're short.
For engineering
Engineering at Luna with AI alongside
Where to put agents in your loop, how to give them context safely, and what's still your call.
5 modules · ~25 minutes total
For clinical
Clinical work at Luna with AI alongside
How to use agents for the parts of clinical and regulatory work that aren't your highest-judgment work — and where the boundary lives.
5 modules · ~25 minutes total
For ops
Operations at Luna with AI alongside
Support, customer ops, vendor coordination, and the long tail of 'who handles X.' Where agents land first, and what stays human.
5 modules · ~25 minutes total
For leadership
Leadership at Luna with AI alongside
Briefings, board prep, investor follow-up, and the meta-question of how leadership uses AI without setting wrong incentives.
5 modules · ~25 minutes total
For admin
AI admin at Luna
What the AI-admin role does, what it explicitly doesn't, and how to handle the parts that affect everyone.
5 modules · ~25 minutes total
Why we have onboarding at all
Because the platform is half technical, half cultural.
The hardest work in shipping an AI workplace isn't the Cloudflare plumbing. It's making sure the team trusts what we built — that the boundaries are real, that the metrics aren't being misread, that "AI assisted" doesn't mean "the human is off the hook." The first module on every track exists for that reason.
Done with onboarding?