Product Manager - Proper OS

Pay or Shift Range

$150,000 USD to $175,000 USD

The estimated range is the budgeted amount for this position. Final offers are based on various factors, including skill set, experience, location, qualifications and other job‑related reasons.

Proper Hospitality is seeking an experienced Product Manager to join our corporate team in Los Angeles.

Hospitality is one of the oldest, most human industries in the world. It has also barely changed operationally in decades. We’re changing that — not with a point tool, not with a clever chatbot, but with a platform that reimagines how an entire property thinks, operates, and serves its guests.

• This is a product, not an internal tool.
• It has to work on day one at a single property and scale cleanly to hundreds.
• Every architectural decision compounds over time.
• Every agent will be onboarded to new properties, new teams, and new data environments.

We are replacing workflows that have been done by hand for decades. Done right, this changes what it means to run a hotel.

You are the translation layer. Department heads tell you what’s broken in their day. You turn that into a spec the engineering team can ship in a sprint. You validate the output. You run the adoption playbook.

You will get your hands dirty. You’ll sit in morning standups at hotels, shadow department heads, build quick prototypes in Claude to test concepts, and write acceptance criteria tight enough that engineering doesn’t have to guess. You will work daily with our engineering lead and the Chief of Staff, CEO, President, and COO.

• Embed with department heads and Hotel GMs to understand their actual workflows — not what they say they do, what they actually do.
• Write agent specs with clear acceptance criteria ("Proposal agent pulls fact sheet + F&B minimums from wiki, generates draft in brand voice, posts to Teams for DOS review. Acceptance: under 15 min from inquiry to draft, DOS edits < 20% of content.")
• Prioritize the backlog. When leadership sends 14 ideas and engineering has capacity for 3, you decide which 3 and defend the decision with data.
• Run the intake process for new agent requests — triage, scope, estimate, schedule.
• Build quick prototypes in Claude or lightweight tools to test concepts before engineering invests real cycles.

• Run the 4‑week adoption playbook per workflow: build with them (Week
  1), validate quietly (Week
  2), soft launch (Week
  3), expand (Week
  4).
• Validate agent output quality before go‑live — is the briefing actually useful? Is the proposal actually in brand voice? Would the GM read this before their standup?
• Track adoption metrics: are people reading the briefing? Has anyone muted the Teams channel? Are department heads referencing agent data in their updates?
• Collect and route feedback to engineering. When a GM says “this alert is noise,” you figure out why and write the threshold adjustment into the next sprint.

• Build working prototypes to test assumptions before committing engineering time.
• Evaluate new AI capabilities weekly — but only deploy what maps to defined operational ROI, not speculative experiments.
• Translate what's happening at the AI frontier into specific, actionable opportunities for luxury hospitality.

• You’ve been a PM for 3–6 years.
• You’ve shipped products that non‑technical users depend on daily — not just internal tools for engineers.
• You understand LLMs and AI agents at a practical level. You’ve used Agentic AI tools extensively (Claude, Openclaw, Perplexity). You built prompts and understand RAG conceptually. You don’t need to write production Type Script, but you can read a spec and spot when something won’t work.
• You’ve sat with end users and changed a roadmap based on what you learned. Not what the exec suggested — what the user actually needed.
• You write clearly. Agent specs, adoption reports, status updates for the CEO — you’re comfortable at every altitude.
• You’re comfortable with ambiguity. This isn’t a Google PM role with a 6‑month planning cycle. It’s a startup PM role where you’re defining the…