01 / The Problem

Nightlife data has a short shelf life, so nobody keeps it.

The useful information lives in group chats, Instagram stories, promoter broadcasts, and word of mouth. It is scattered, unsourced, and expires within minutes. People still arrive to the wrong music, an unexpected cover, a dead room, no dancefloor, or nowhere to park.

Existing discovery platforms describe places well. They describe the state of a place badly — and state is the only fact that matters when you are standing outside on a sidewalk deciding which door to enter.

OUTBND is built around a single design doctrine: Wrong data is worse than no data. Freshness is a strict computational constraint rather than a visual styling option. A venue whose status has aged out never renders a stale genre or cover price as though it were current — it drops the reading and displays —.

02 / Interactive Engine

Inspect the live map, 4-level energy gauge, and detent sheet.

Simulate live updates and watch the client state machine handle real-time WebSocket frames, energy radiation, and honest freshness degradation.

Select Pilot Venue Montrose / Midtown

Simulate Recency & Age Freshness Ladder

Energy Gauge Levels Level + Color

Quiet #7C6A5F
Warming #C4855A
Busy #FF7D55
Packed #FF4D26
No Live Data #55524A

iOS Sheet Detent 3-Level Detent

WebSocket Relay Realtime Push

11:14 PM OUTBND LIVE

Numbers Nightclub

Montrose · Houston

LIVE · 3m ago

Energy Level Packed energy

Door Cover $15

Music Genre Industrial / Darkwave

Walking Distance 8 min walk (0.4 mi)

Dancefloor Large Hall + Stage

Parking Rear Lot & Street

Signal Source Venue Staff Tonight Card

03 / Signal Architecture

Four feeds, three live reporters.

Verified venue facts supply durable ground truth. Live room state flows through three real-time reporting channels: the people running the room, the crowd inside it, or ambient edge sensing.

OUTBND_SIGNAL_PIPELINE.DECK // 4 INGESTION CHANNELS

FEED 01 / 04

Channel 01 · First-Party Broadcast

Venue & Staff Updates

Direct Attribution · Highest Trust

A streamlined, authenticated "Tonight" publisher built directly for venue operators, managers, and door staff. Enables immediate, authoritative broadcast of current room genre, energy level, cover charges, line velocity, and door conditions.

Input Vector: iOS & Web Door Publisher
Confidence Rating: 98% Direct Verification
Expected Cadence: 1–4 updates / night
Primary Risk: Staff omission during peak midnight rush

Implementation Status Production Ready — FastAPI Tonight API & iOS Editor Deployed

Channel 02 · In-Venue Consensus

Patrons in the Room

High Volume · Spatial Consensus

Rapid observational state reports submitted exclusively by patrons whose physical presence within the venue geofence is cryptographically and spatially verified. Crucial for live coverage across independent venues lacking active staff publishers.

Input Vector: Geofenced Patron Observations
Confidence Rating: 74% Crowd Consensus (≥ 3 reports)
Expected Cadence: 15–60 observations / night
Primary Risk: Promoter gaming & location spoofing

Implementation Status Phase 2 Architecture — Spatial Consensus & Anti-Sybil Clustering Pipeline

Channel 03 · Ambient Edge Telemetry

SAMI

Physical Readings · Zero Human Friction

SAMI is OUTBND's ambient signal layer. It contributes continuous room-state confidence when direct staff or patron updates are sparse, while preserving the platform's privacy-first posture.

Input Vector: SAMI Ambient Room-State Feed
Confidence Rating: 92% Deterministic Telemetry
Expected Cadence: Continuous 60-second frames
Primary Risk: SAMI signal temporarily unavailable in a venue

Implementation Status Active deployment channel with confidence-weighted ingestion.

Channel 04 · Canonical Baseline

Verified Venue Facts

100% Deterministic · Spatial Baseline

The immutable spatial foundation: verified physical addresses, polygon boundaries, standard operating hours, cover fee conventions, parking logistics, and structural entrance rules managed in Postgres and PostGIS spatial tables.

Input Vector: Curated Spatial DB & Verified Schema
Confidence Rating: 100% Ground Truth
Expected Cadence: Static / Weekly Administrative Verification
Primary Risk: Unannounced holiday closures or format shifts

Implementation Status Live & Deployed — Supabase Postgres + PostGIS Spatial Engine Active

04 / The Trust Model

A missing reading beats a confidently wrong one.

Every live claim carries a source class and an exact timestamp. Stale claims are removed from the decision surface rather than shown faintly. Conflicting reports are displayed as disagreements instead of averaged into false certainty.

Adjust Age Since Last Broadcast: 3 min

TIER 1: LIVE (<15m)

Genre Darkwave / Techno

Cover $15

Freshness Decay Rule

High confidence. Displayed with full hue prominence and active radar radiation.

05 / SAMI

SAMI extends live room-state confidence.

SAMI is the ambient signal layer in OUTBND. It contributes confidence-weighted room-state input between human updates and keeps the system honest when direct reporting pauses.

One role: support continuity.

SAMI supports venue-state continuity as a supplemental feed, not a hidden authority. Its contribution is always source-attributed and interpreted through the same decay rules as every other live claim.

01 / Passive Ambient Input

SAMI supplies low-friction ambient signals for room-state confidence without requiring customer interaction.

02 / Confidence-Weighted Ingestion

SAMI readings are fused as one channel in the trust model and never override verified venue facts or direct updates.

03 / Honest Degradation

If SAMI signal quality drops, OUTBND degrades visibility instead of showing stale confidence.

SAMI / Signal Channel 03

Ambient signal in, confidence out.

Role in the trust model SAMI helps bridge reporting gaps between direct human updates and verified baseline facts.

If SAMI confidence drops or signal continuity is interrupted, OUTBND reduces claim visibility and never leaves stale certainty on screen.

06 / Technical Architecture

Built on fast, honest foundations.

The entire stack is arranged around one question: How fast can a phone find out what is happening near it right now, and how confident should it be in the answer?

Layer	Implementation	Architectural Rationale
iPhone Client iOS 17+ · Native	SwiftUI + Custom Mapbox Style + MapKit	One screen, one-handed in the dark. Mapbox delivers custom nocturnal vectors; MapKit handles turn-by-turn walking times.
Backend API FastAPI · Python Async	Deployed on Fly.io (Two Always-On Nodes)	Permanent WebSocket connections relay status changes within milliseconds. Machines stay awake to prevent connection drops.
Spatial Database Postgres + PostGIS	Hosted on Supabase (Database Only)	PostGIS computes "nearest open venues within 3km" natively. Row-Level Security enabled; Supabase Auth/Realtime bypassed.
Realtime Sync Redis + WebSockets	In-memory event publication broker	When a venue's status updates, connected clients update rows in-place without polling or interrupting gestures.
Contract Safety OpenAPI Generation	Automated Swift codegen scripts	Swift models are generated directly from the backend schema, ensuring client and server never diverge on data structure.
Staff Authentication Cloudflare Email Sending	Pre-provisioned verified staff login	Staff activate by secure emailed code and salted hash. Prevents unauthorized claims while keeping customer sign-in passwordless.

07 / Rollout & Model

Density before geography.

One nightlife district made genuinely useful, rather than a scatter of unreliable dots across a city. The pilot focuses on Houston's core nocturnal corridors.

Principle 01

Consumer Access is Free

Patrons access live maps, energy levels, music previews, and walking ETAs without paywalls or subscription barriers.

Principle 02

Operator Tools & Verification

Monetization centers on venue-side management tools, verified business claims, and distinct promotional disclosures.

Principle 03

Aggregated Scene Analytics

Anonymized, district-wide nightlife foot-traffic trends and genre preferences sold back to operators for event planning.

“The consumer product works. Turning a disclosed demo into a trusted live nightlife network is what is left, and it is the part worth doing.”

Where nightlife changes by the minute, the map has to tell the truth.