Infrastructure for collective intelligence.
We build the infrastructure that turns team communication into permanent organizational wisdom—for any team where knowledge compounds and relationships define quality. No apps, no data entry, nothing to log into.
The Most Valuable Knowledge Lives in People. And People Leave.
Every organization has people who know things no one else knows. The senior account manager who remembers which clients need hand-holding. The engineer who knows why that system was built that way. The nurse who can tell when a patient is declining before the vitals show it. When these people leave, the knowledge leaves with them.
In hospitality, this problem is acute. A ryokan apprentice in Kyoto learns through years of quiet observation—how to read a guest's mood from the way they remove their shoes, which room to offer a couple celebrating an anniversary, when to appear and when to disappear. This knowledge is never written down. It lives in the body, in instinct, in accumulated attention.
Andrew spent 12 years at the front desk of The Sycamore. He knew that Mr. Chen always arrives on red-eyes and needs a quiet room away from the elevator. That the Hendersons celebrate their anniversary every March and prefer the fourth floor. That Mrs. Reyes will say everything is fine but mention problems in her review if you don't ask twice.
When Andrew left, none of this went with the hotel. It walked out the door with him.
This isn't a hospitality problem—it's a structural one. The most valuable organizational knowledge—pattern recognition, relationship history, situational judgment—has never had a place to live outside of individual memory.
Soul, Not Robots
AI is splitting all work into two categories. In some jobs, humans are the means to an end—those get automated. In others, humans are the product: the nurse who reads a patient's fear, the advisor who knows when to push back, the concierge who makes you feel known. These jobs don't need automation. They need augmentation.
The default response to the knowledge problem is automation: deploy chatbots, script interactions, replace human judgment with algorithms. This misses the point entirely. The best work isn't efficient—it's attentive. You can't automate the instinct that tells a nurse when a patient needs attention, or the judgment that tells an account manager when to push back versus when to accommodate. In hospitality, you can't automate the instinct that tells Maria to comp breakfast before a guest complains.
- → Chatbots that handle people "so staff don't have to"
- → Automation that replaces judgment with scripts
- → "Efficiency" that erodes what makes service human
- → Collective intelligence that amplifies your team
- → Memory that persists when people leave
- → Judgment that compounds, not depletes
Omote doesn't automate care. It makes the people who provide care collectively smarter.
Augmentation over automation.
The default assumption in AI is replacement—remove humans from the loop, optimize them away. We reject that framing. The best service will always be human. AI's job is to make those humans extraordinary, not expendable. This isn't just a product philosophy. It's a bet on how this technology should be applied.
Omote captures context from the channels your team already uses—voice conversations, WhatsApp messages, forwarded emails, photos of guest notes. No logging, no typing, no new workflow. The system learns from how your team actually communicates, and feeds that wisdom back as real-time guidance.
When Andrew leaves, what he knew stays. When a new hire starts, they inherit the instincts of every operator who came before them.
Three Primitives Turn Communication Into Collective Intelligence
Grove, Stream, and Stone aren't hospitality features—they're infrastructure primitives grounded in universal concepts: memory, awareness, judgment. The nature metaphors are intentional. A Grove grows in any domain. A Stream flows through any operation. A Stone accumulates anywhere humans learn by doing.
Hotel: "Mr. Chen — prefers high floor, shellfish allergy, always on red-eyes."
Care: "Mrs. Johnson — daughter visits Sundays, prefers morning check-ins, hearing aid in left ear."
Hotel: "Room 412 noise complaint → Maria responding → room move accepted → follow-up 9am."
Care: "Mr. Davis fall risk flagged → PT scheduled 2pm → family notified → mobility aid ordered."
Hotel: "Late-night noise? Offer room move first—94% success. Comp breakfast if they accept."
Care: "When Mrs. Johnson seems withdrawn, check if daughter's visit was cancelled. Usually is."
In Practice: Hospitality Example
The same pattern applies to any domain where relationships matter. Here's how it plays out at a hotel:
Mr. Chen arrives at midnight after a delayed flight. A new hire is at the desk—it's her second week. She doesn't know Mr. Chen. But as he approaches, her earpiece chimes:
Grove remembers Mr. Chen — 4 stays, always on red-eyes, prefers high floor away from elevator, shellfish allergy. His birthday is next week.
Stream shows his flight was 3 hours delayed. Room 814 is ready, pre-set to 68°F. Kitchen closes in 20 minutes—late menu available.
Stone suggests: Delayed red-eye guests appreciate a quiet check-in. Offer the late menu before he asks. Mention his birthday—personal touches score 89% positive response.
She greets him by name before he reaches the desk. Confirms his quiet room on a high floor, mentions the kitchen is still open if he'd like something sent up, and wishes him an early happy birthday.
Mr. Chen pauses. "How did you know all that?"
She didn't learn this over years. She inherited twelve years of institutional wisdom in twelve seconds. That's collective intelligence.
Nothing to Log Into
Omote has no dashboard.
Most enterprise software is a screen workers are supposed to check between actual work. A tab to keep open. A system to learn. Another place to enter data. Omote has none of that.
Talk to Omote like you'd talk to your best colleague. WhatsApp, voice, SMS—whatever channel your team already uses.
"How did last night go?"
"Quiet night. 94% occupancy. Two noise complaints resolved. Hendersons asked about late checkout—flagged for morning."
Need a report? Ask for it. Charts, summaries, and analyses generated on demand—export to PDF to share with ownership.
"Create a guest satisfaction report for January for the owners."
→ PDF generated with trends, complaint patterns, and recommendations. Ready to send.
Context flows in however it's natural. Photos, voice, email, screenshots—all parsed automatically.
Photo
Guest note → profile
Booking → parsed
Voice
Mention → noted
Screenshot
PMS → extracted
Every Conversation Makes the System Smarter
Omote learns continuously from how your team works. Not from training sessions—from real operations.
Operators work
Context flows through Stream
Stone suggests
Patterns surface as guidance
Team responds
Accept, modify, or reject
Stone learns
Patterns refined by outcome
Accepted
Pattern reinforced
Modified
Pattern refined
Rejected
Pattern questioned
Outcome
Success or failure
A property that's used Omote for a year has a Stone that reflects thousands of real decisions. Patterns start as suggestions, gain confidence through use, and become institutional playbooks. A new hire on day one inherits all of it.
Day Zero Isn't Empty
Most AI systems start blank and need months of data before they're useful. Omote doesn't.
Omote ships with a foundational layer grounded in Japanese Omotenashi—the philosophy of anticipatory, selfless service. Concrete decision patterns: resolution paths, timing heuristics, judgment frameworks. Not generic advice—operational wisdom from the world's best properties.
During setup, the Omote team embeds with your property to extract the tacit knowledge that lives in your team's heads:
- • Structured interviews — "Walk me through a late-night noise complaint"
- • Shift observation — shadowing operators to capture real workflows
- • Artifact collection — cheat sheets, handoff notes, the binder behind the desk
- • Action tracing — mapping decision sequences, not just outcomes
Omote is useful on the first shift.
The Omotenashi base provides the foundation. Your property's own knowledge layers on top. Live operations refine it daily.
Properties own their data—their Grove, their Stream. But the anonymized patterns that emerge across properties? How the best teams in the world handle noise complaints, VIP arrivals, service recovery? That's Omote's proprietary asset. Every new property makes the system smarter for all of them.
This is the data moat: not raw conversation logs, but distilled operational wisdom no competitor can replicate from scratch.
This Wasn't Possible Two Years Ago
The technical capabilities that enable zero-friction knowledge capture have only recently matured:
Now fast, accurate, and cheap. A 30-second conversation transcribes in under a second and extracts structured context—guest names, room numbers, preferences, requests—without templates or parsing rules.
Any system becomes "integrated" with a photo. A screenshot of a PMS, a photo of a handwritten note, an email confirmation—AI extracts structured data instantly. No OCR pipelines, no API partnerships.
Normalize chaos. Paste a messy email, get structured guest data. Forward a booking confirmation in any format, from any source—Omote makes sense of it. No rigid integration required.
Integration no longer means APIs, partnerships, and months of IT work. It means: "Send us whatever you have. We'll figure it out."
Voice-first wearables are coming.
Apple, Meta, OpenAI, and others are building the hardware. Omote doesn't compete with them—we're the collective intelligence layer that makes any device smarter. As the hardware surface area expands, our capture advantage compounds.
The Urgency
The hospitality industry faces a 43 million worker shortfall by 2035. This isn't a future problem:
The crisis isn't just labor scarcity. It's knowledge scarcity. Every departure is a loss of institutional memory. Every new hire starts from zero. The gap between what experienced operators know and what new operators can learn is widening every year.
The technology to capture tacit knowledge just became available. The workforce crisis makes it urgent. The window is now.
Hospitality Is the Wedge, Not the Destination
We start with hospitality because the conditions are perfect: acute pain, clear ROI, fast feedback loops. But the architecture—connecting frontline humans to collective intelligence through memory, awareness, and judgment—applies wherever the human relationship defines quality.
Where the pain is sharpest: 74% annual turnover, knowledge that directly drives revenue, and buyers with budget authority. Every hotel knows exactly what was lost when their best people left.
The conditions are perfect: acute pain, clear ROI, fast feedback loops. Hospitality is the proving ground. The infrastructure is built to travel.
Senior living where residents depend on being known. Healthcare where provider handoffs lose critical context. Wealth management where relationships span decades. Legal services where matter history drives outcomes.
Grove, Stream, and Stone aren't hospitality concepts—they're infrastructure primitives for any domain where knowledge compounds and relationships define quality.
Customer success teams. Field service organizations. Consulting firms. Any team where "the person who knows" is a bottleneck and institutional memory creates competitive advantage.
Self-serve onboarding. Domain-specific patterns as add-ons. Collective intelligence as infrastructure.
The Team
The team combines product experience (Google, Sonder), business modeling (McKinsey, Harvard MBA), and direct experience building and operating hospitality properties.
We believe the best service will always be human. AI's job is to make humans extraordinary—not to replace them.
Let's Build This Together
Omote is in active development, building alongside teams who believe their collective knowledge deserves to be permanent.
If you run a team where memory and relationships are the product—or if you invest in the future of knowledge work—we'd like to talk.
Reading this from outside hospitality and thinking "this would work for my team"? You're probably right. Let's talk.