Ramada

Live in production

Tell it what you're in the mood for tonight in Tucson, and it ranks places against simply staying home.

Ramada showing staying in ranked above the best places to go out tonight

Current status

Live Tucson night-out planner running in the browser at ramada.ngi.video with no account required.

Overview

A Tucson night-out planner that checks the clock, live weather, drive time, and the mood you pick, then ranks the best places to go tonight against simply staying home.

Why it was built

Map apps and event sites dump hundreds of pins on a screen without answering the real question: 'Given the weather, drive time, and my energy right now, what is actually worth leaving the house for?'

Why it helps

Lets you pick a mood, a time, and how long you have, shows clear pros and cons for every venue or trail, and honestly tells you to stay in when nothing beats a night at home.

What you can do

Compares local venues, trails, libraries, and events against a realistic staying-home baseline
Mood choices: Anything, Something to do, Meet people, Be around people, Get things done, or Unwind
Real-time weather, daylight window, and drive-time awareness for tonight or the weekend
Transparent preference controls so you can see and adjust why any spot was recommended

Ramada • Core Technology & Feature Architecture

Epistemic Field Decision Engine, Sonoran WBGT Physics & Stay-Home VOI Baseline

Models every Tucson venue and trail attribute as an epistemic Field { mean, confidence, source, n } and benchmarks every night-out itinerary against a calibrated 0.58 Stay-Home baseline.

Epistemic Field Invariants

Strict absent() vs. prior() & Anti-Silent-Exclusion Gate

Never collapses missing data into priors; hard filters only exclude a venue when field confidence >= 0.60, keeping low-confidence spots visible and badged 'unverified — scout' instead of silently hiding them.

Stay-Home VOI Gate

Honest 0.58 Stay-At-Home Baseline Across 6 Moods

Evaluates 8 utility estimators across 6 user moods against homeValuePerStayHFloor (0.58) so the planner honestly ranks Stay Home #1 when heat, monsoon rain, or drive time make going out net-negative.

Commit vs. Explore Rank

Conservative Lower-Bound (score - k·spread) & Scout Mode

Sorts Tonight commitments by conservative lower bounds (penalizing wide uncertainty intervals) while Explore/Scout mode sorts by upper bounds (score + k·spread) to surface high-upside unverified spots.

Sonoran Physics Engine

Tucson Wet-Bulb Globe Temp (WBGT), Twilight & Travel Math

Computes live desert heat stress (WBGT), solar twilight windows, Shenandoah trail difficulty (1–5), and multimodal travel friction inside a pure zero-I/O functional core.

Ramada • Epistemic `Field` Interval & 0.58 Stay-Home Baseline Lab

Test when Tucson heat, drive time, and field uncertainty make Staying Home rank #1

Tucson evening temperature (`wbgt.ts`):98°F

One-way drive time (`travel-cost.ts`):18 min

Venue attribute confidence (`minConfidenceToExclude = 0.60`):0.45

Tonight Recommendation Verdict

Venue commit lower-bound (score - k·spread): 0.46 (mean 0.64, spread ±0.18)

Stay-Home VOI baseline (homeValuePerStayHFloor): 0.58

Filter exclusion gate: Kept in list & badged "unverified — scout" (conf < 0.60)

Top-ranked Tonight choice: 1. Stay Home (0.58 beats going out in current conditions)

Before & After Engineering AuditEpistemic Activity & Night-Out Decision Engine0.58 Stay-Home honesty gate · minConfidenceToExclude = 0.60

Epistemic Field intervals, Tucson WBGT & 0.58 Stay-Home baseline vs. map pin dumps

Comparison focus

Before: 200 static '4.5-star' pins ignoring 104°F heat, twilight & thin reviews

Map apps recommend outdoor trails at 104°F or cafes closing in 30 minutes, and silently hide venues when a single filter attribute is unknown.

// Standard map filter
venues.filter(v => v.rating >= 4.5 && v.hasOutlets === true)
// Bug 1: Silently drops great cafe where hasOutlets is unknown (null)
// Bug 2: Recommends hike at 103°F WBGT with 18 mins of daylight left

After: Field { mean, confidence } intervals, 'unverified — scout' badge & Stay-Home gate

Ramada models every attribute as an epistemic Field, keeps low-confidence venues visible with an 'unverified — scout' badge (minConfidenceToExclude = 0.6), and ranks Stay Home (0.58) #1 when going out loses.

// Ramada Pure Epistemic Engine (packages/engine)
outletField = { mean: 0.5, confidence: 0.25, source: "absent" } // < 0.60 -> kept as "unverified — scout"
commitScore = score - k * spread // conservative lower-bound rank
if (bestVenue.commitScore < stayHomeBaseline(0.58)) return rankFirst("Stay Home")

CV & VLM rubric shift: 0.776 → 0.9525 (+0.1765 · Sharpness +50.8% · PSNR 44.81 dB)

Grounding gate: Negation parity & evidence verified

Interactive walkthrough

Step through how Ramada works without creating an account or installing local dependencies.

Entry & problem context

Maturity: Live in production

Map apps and event sites dump hundreds of pins on a screen without answering the real question: 'Given the weather, drive time, and my energy right now, what is actually worth leaving the house for?'

A Tucson night-out planner that checks the clock, live weather, drive time, and the mood you pick, then ranks the best places to go tonight against simply staying home.

Shared platform architecture sandbox

Test the live cross-language single source of truth shared across TypeScript, Python, and Rust.

Normalize legacy or rolling model slug

Resolved OpenRouter wire slug: ~anthropic/claude-sonnet-latest

Contextual task intent routerPrefer zero-cost floor (`openrouter/free` cap enforced)

Primary wire model: google/gemini-3.1-flash-image (Gemini 3.1 Flash Image)

Capped fallback chain: google/gemini-3.1-flash-image → google/gemini-nano-banana-2.1 → openrouter/free

Pareto composite score: 0.8499 (effective cost $0.00102/1k)

Universal 9-Surface Visual & Image Capability Router (`resolveVisualRoute`)

Routes character turnaround sheets, 3:4 VN busts, 16:9 scenes, top-down battlemaps, 1:1 tokens, typography plates, semiotic brand favicons, WebGPU background removal, and Tier-2 local ComfyUI models.

Prefer OpenRouter image gateway

Primary model: black-forest-labs/flux-2-pro via fal

Aspect ratio: 16:9 · Max ref images: 2 · Fallback: google/gemini-3-pro-image → google/gemini-3.1-flash-image

Cinematic 16:9 scene backdrop and storyboard synthesis (`fableforge`, `containment`, `ngi.video/console`).

Anchor promotion (`score >= 0.95`): Promoted to canonical anchor (overall 0.958)

Speculative YOLO anatomy pass: Standby · Vector guard: 1024d verified

DeepSeek 18:00 UTC window: Off-peak active (50% V3 / 25% R1 cost)

Model-aware prompt compiler & five-axis vision rubric (30 registered models)

Compiled prompt (natural_language):

Noir detective standing under neon rain in a cyberpunk alley, cinematic volumetric lighting Character anchors: trenchcoat; amber cybernetic eye. Composition & pose guidance: 16:9 widescreen; cinematic keyframes.

Gemini Vision five-axis rubric weighted score: 0.9 (Passes quality gate)

Semiotic Brand Identity & Federated SoC Governance

Peircean Brand Mark, Multi-Size Squint Test & 3-Tier SoC Boundary

SoC tier: internal_commercial16px rubric: 96.7% (PASS)

16px · 32px · 48px squint test

Motif: decision_tree_scale

Iconic sign: Bifurcating decision-tree nodes balanced on a precision fulcrum

Indexical cue: Expected-value (EV) utility calibration and Fuzzy-Trace gist synthesis

Symbolic archetype: Teal-cyan epistemic clarity and rational choice compass

SoC Boundary & Epistemic Retrieval Guardrails

Visibility & ceiling: private_commercial (max tier: internal_commercial)

Boundary policy: Commercial hospitality & leisure epistemic decision engine.

4-gram shingle novelty: 97.7% (non-redundant >15%) · Retired symbol trap: Clear (0 retired symbols)

GitHub Profile & Source Activity

Engineering Velocity & Open-Source Hygiene

Synced from @brbrainerd and @vox-foundation via the GitHub GraphQL API (GET /api/github/stats).

GitHub Profile vox-foundation/vox

Annual GitHub activity

13,842+

3,229 public · 10,136 private revisions

Curated portfolio history

19,523+

Revisions across 32 curated projects

Pull requests & reviews

463 PRs

Merged across 70 first-party codebases

Ramada historyTypeScript

1,121 revisions

1,884 tracked source files · Updated 2026-10-18

Language distribution across curated projectsWeighted by revision history & source volume

TypeScript38.9% (12 projects)

Rust36.9% (1 project)

Python11.3% (8 projects)

JavaScript4.7% (2 projects)

Lua4.1% (1 project)

Public open-source codebases on GitHub

vox-foundation/voxRust · 7,200 rev

Single-file full-stack programming language and Rust compiler that generates database schemas, server routes, and typed clients from one declaration.

#programming-language#rust#compiler#full-stack#lsp

brbrainerd/govSimTypeScript · 435 rev

Interactive political economy simulator with ECS civic rights, institutional modeling, policy levers, and multi-regime scenarios.

#simulation#political-economy#typescript#ecs#strategy-game

brbrainerd/neurogeckoHTML · 6 rev

Interactive 3D macaque brain atlas and stereo surgical view planner rendered in browser WebGL.

#neuroscience#webgl#threejs#brain-atlas#3d-visualization

brbrainerd/brFrameTypeScript · 23 rev

Single-binary Rust and React kiosk shell for Raspberry Pi digital photo frames with remote mobile upload.

#rust#raspberry-pi#kiosk#embedded#react

Open Ramada Browse all projects