Selected Work / 2026, Hospitality
Tacet: The
Environmental
Twin.
A that turns street, weather and crowd signals into yield rules your RMS can act on, exposed via .
Where this sits
A food & beverage execution agent for hotels. It tells a manager how many covers to expect, how to staff for them, and how to cut food waste.
- For
- Other agents via MCP, or a hotel / F&B manager with WhatsApp.
- Status
- Built and instrumented, pilot partner in recruitment.
- TacetYou are here
Headless environmental twin: street, weather and crowd signals turned into yield rules, exposed via MCP.
- For
- RMS, PMS and orchestration agents, not humans.
- Status
- Built: live data ingestion pipeline, MCP server v0 live on Fly.io.
A voice AI mentor that captures, qualifies, and leverages the tacit expertise from senior to junior collaborators. Just ask.
- For
- A junior technician on the shop floor, hands busy.
- Status
- Prototype, synthetic data, harness in CI.
Guest node: memory that makes every guest a regular, anticipating needs and learning from every stay without watching.
- For
- The nodes that need to know who is arriving.
- Status
- Synthetic PoC, production gated behind a data-protection assessment.
Open-data SaaS mapping urban noise so cities can act on acoustic health.
- For
- European cities and metropolitan areas.
- Status
- Shipped 2020 to 2022, company closed.
Project Info
Rôle
Product Builder, Architect of an environmental intelligence layer
Durée
2026, ongoing (V3, MCP server and Dual Memory)
Équipe
Solo (with AI collaboration)
Contexte
Headless Environmental Twin for hospitality, exposed via MCP, complement to Aetherix
Stack
Python 3.12, FastAPI, Anthropic MCP SDK, OSMnx and Shapely ray-tracing, SQLAlchemy dual-memory store, Mews and Apaleo connectors
Repository
GitHub RepositoryFrom dashboard
to anticipation
Tacet began as an exploration of urban noise pollution, a Bruitparif-powered Score Sérénité dashboard for Parisians.
The true value of data, however, lies not in visualisation but in anticipation. Tacet became a : a physics-based model of the property and its surroundings, exposed via an server that any RMS, PMS or orchestration agent can consume.
RMS and PMS run on history.
A transport strike, an unannounced street excavation, a sudden heatwave: the street, the sky and the crowd are not in the booking data. That blindness shows up as mispriced inventory, operational friction and avoidable refunds.
Environmental twin, isometric view
A physics-based, software-only model of the property and its surroundings.
An environmental twin, exposed via MCP
Instead of pushing yet another dashboard, Tacet publishes calibrated yield rules and a mathematical on an server. Sense, reason, expose.
Sense
The engine continuously ingests proxy APIs: construction permits, traffic flow, weather conditions, local events and mobility disruptions.
Reason
A 3D ray-tracing physics engine, paired with a dual-memory store, turns raw signals into a calibrated impact score for each room of each property.
Expose
Yield rules and a mathematical Chain of Thought are published on an MCP server, ready for the RMS, the PMS, or an orchestration agent like Aetherix.
"Heatwave forecast plus rail strike on June 13. Tacet suggests minus 15% on street-facing suites for two nights, and a staffing buffer for breakfast service. Awaiting Revenue Manager approval."
The intelligence pipeline
From proxy signals to MCP rules, the headless flow that turns context into decisions.
OpenData Permits
Webhooks
Traffic APIs
Streams
Weather APIs
Streams
Local Events
Streams
Ray-tracing Physics Engine
3D line of sight, shielding
Dual Memory store
Idiosyncratic + Hive Mind
Revenue Management
Yield rules, HITL approval
Property Management
Room tags, reassignment
MCP server: rules and Chain of Thought
Consumed by RMS, PMS and orchestration agents
Proxy signals · Ray-traced physics · Dual memory · MCP rules
Four pillars give Tacet its edge
Physics, memory, restraint and explainability. Each pillar protects the others against the failure modes of a headless agent.
Spatial Physics Engine
Tacet does not just measure distance, it measures physics. Using shapely and osmnx, it draws 3D lines of sight between a disruptive event and the property, then applies a shielding penalty when urban buildings block the path. False positives drop, alert fatigue disappears.
Dual Memory
Idiosyncratic memory learns each property's quirks: a hotel that rejects "traffic noise" alerts probably has triple-glazed windows, so Tacet applies a persistent shielding bonus at that GPS point. Hive Mind aggregates rejection rates across the network to recalibrate baseline sensitivity.
Human-in-the-Loop Guardrail
Tacet never executes destructive actions on its own. It formulates a precise Price Modifier Rule (for example, minus 15% on street-facing suites between June 12 and 14) and pushes it to the RMS for a Revenue Manager to approve. Accountability stays with the team.
Conversational Explainability
Every JSON payload carries a mathematical Chain of Thought: base noise, distance attenuation, ray-tracing penalty, final impact. An orchestration agent parses it and explains the decision to the Revenue Manager in natural language, without a GUI.
Impact, beyond yield
Embedded in an agentic OS, Tacet shifts properties from reactive to proactive. Yield protected, operations smoothed, footprint reduced.
Yield protection
Pricing noisy inventory ahead of time prevents costly post-stay refunds and protects average daily rate on the rest of the building.
Operational excellence
When a heatwave overlaps with a transport strike, the orchestration layer adjusts F&B prep and staffing buffers before the day starts.
Sustainability
Less energy waste through proactive HVAC optimisation, less urban noise pollution carried into rooms, measured against a stable baseline.
Tacet and Aetherix
Tacet exposes its rules and reasoning through . Aetherix consumes them as one more tool in its agentic mesh, forming a complete predictive twin of the property.
Part of the Hospitality Agentic Mesh: the architecture note explains how the five agents divide the work.
Aetherix, internal context
Anticipates operational needs, F&B demand and staffing, based on occupancy and historical trends.
Tacet, external context
Reads street, weather and event signals, ray-traces their impact, and publishes yield rules on MCP.
Together, they let teams control costs, reduce waste and protect the experience guests came for.
Where Tacet is going
Widen the catalogue of signals (mobility, air quality, cultural events), invite more orchestration agents on the MCP layer, and let the dual memory get sharper at each property.
Over time, Tacet becomes the environmental layer of an agentic OS for hospitality: each property reads its surroundings with its own context, and responds faster because of it. A Quiet Certification standard is a likely side-effect, not the goal.
Pilot Tacet with your hotel
Looking for design partners among boutique hotels in Europe.
Frequently Asked Questions
Why headless and exposed via MCP?
MCP is becoming the standard contract between agents and tools. Tacet ships as an MCP server so any RMS, PMS or orchestration agent can read its rules and reasoning without a bespoke integration.
How accurate is a software-only model versus rooftop sensors?
Tacet ray-traces 3D building geometry from OpenStreetMap and calibrates against public acoustic baselines like Bruitparif. No capex, no hardware to maintain, and the model gets sharper as the dual-memory store learns each property.
Does Tacet change prices on its own?
No. Tacet formulates a Price Modifier Rule with an explicit window and inventory scope, then sends it to the RMS for a human Revenue Manager to approve. Accountability and brand control stay with the team.
How does Tacet relate to Aetherix?
Aetherix reads the internal context (occupancy, F&B, staffing). Tacet reads the external context (street, weather, events) and exposes it through MCP. Together they form one predictive twin of the property.