TOGA — User Architecture

Workflows • Decision Engine • Data Architecture

← BACK TO SITE
01

Onboarding

One-time, 10-15 minutes. Conversational, not a form.

1Identity
2 min
  • Base (IAH, ORD, SFO…)
  • Equipment (737, 777, 787…)
  • Position (CA / FO)
  • Seniority (system + category)
  • Employee file number
  • Commuter? Airports + preferred flights

Category, G-Line position, trip pool

2Lifestyle
3 min
  • Default mode: Days off / Income / Balanced
  • Family: partner, kids (ages), events
  • Recurring constraints (Tuesdays for soccer)
  • Risk tolerance: Conservative → Aggressive
  • SO access: invite partner

Layer 5 (Lifestyle) + Layer 7 (Memory)

3Preferences
3 min
  • Layover favorites + deal-breakers
  • Trip type: turns vs multi-day, dom vs intl
  • Time of day: morning / night / red-eye tolerance
  • Trip length preference (1-4 day)
  • No-fly list (crewmembers to avoid)
  • Existing bidding strategies

Layer 7 (Individual Memory)

4Data Handoff
2 min
  • Current schedule (CalDAV or screenshot)
  • Recent bid award report (PDF)
  • Seniority report
  • LPA/tolerance parameters (PDF)
  • PBS Home screenshot
  • Standing bid description

Baseline for blind spot analysis

5Calendar Sync
2 min
OPTIONAL
  • Connect Google / Apple Calendar
  • Share family calendar for constraints
  • Notification preferences for SO
  • Key family dates for the year

Calendar sync + SO Access features

AFTER ONBOARDING →

TOGA immediately runs the Blind Spot Analysis on the pilot's standing bid and delivers the first Copilot Briefing — showing what they're leaving on the table.

02

Monthly Cycle

The core recurring workflow. Six stages per month.

15 min

Monthly Check-In

Voice / Chat
  • Vibe: Chill / Money / Balanced
  • Hard constraints (must-have days off)
  • Soft preferences (SAN, 3-day trips)
  • Special: no red-eyes, avoid Captain X
  • Risk tolerance for this month
210 min

Blind Spot Review

Chat / Briefing
  • Unreachable preferences (100% senior capture)
  • Hidden tradeoffs (SAN = red-eye legs)
  • Ineffective constraints (LIMIT ≠ BAN)
  • BG cascade gaps (is your floor strong?)
  • Combi bid opportunities
  • Credit optimization review
  • Seniority context + peer comparison
320–60 min

Bid Entry

In PBS GUI
  • Tier 1: Standing bid + tweaks (20 min)
  • Tier 2: Voice walkthrough (30-45 min)
  • Tier 3: Full build (45-60 min)
45 min

Award + Trade Plan

Review
  • Score awarded schedule vs goals
  • Identify what worked / what didn't
  • Surface trade opportunities
  • Premium pay potential
  • Generate trade plan (3-5 moves)
5Ongoing

Active Month

Autonomous
  • Monitor pool display (staffing levels)
  • Track open time (new trips)
  • Alert on premium pay opportunities
  • Detect no-fly list conflicts
  • Auto-sync family calendar
  • Execute approved trades
62 min

Post-Month Review

ROI Card
  • Days off gained (from trades)
  • Premium pay earned
  • Shape improvement score
  • Trade success rate
  • TOGA ROI vs subscription
  • Pilot rates month 1-10
  • Stored in memory → future optimization

Click any stage to expand details.

03

Decision Engine

7-layer hierarchy. Broad → narrow. Each layer refines the one above.

L1
UniversalEvery United pilot
L2
CategoryBase + Equipment + Position
L3
Seniority BandJunior / Mid / Senior
L4
G-LineLineholder vs Reserve
L5
LifestylePersonal circumstances
L6
Monthly VibeThis month's goals
L7
Individual MemoryLearned over time
BLIND SPOT FILTER

Sits between the decision engine output and bid generation. TOGA interrogates its own recommendations before presenting them.

1.What trips in the pool match this preference?
2.What's the senior capture rate?
3.What hidden characteristics do matching trips have?
4.Does this preference conflict with another?
5.Is there a more reachable alternative?
OUTPUT → Copilot Briefing → Pilot decisions → Final bid
04

Interface Modes

TOGA operates across multiple surfaces depending on context.

🎙

Voice

Check-in + Entry
  • Monthly check-in while driving
  • Bid entry walkthrough (AirPods + PBS)
  • Mid-month adjustments on the go
  • 'Hey TOGA, how's my month looking?'
💬

Chat

Strategy + Detail
  • Copilot Briefing (blind spot review)
  • Deep dives on preferences
  • 'Why did TOGA suggest this?'
  • Trade plan review & approval
📊

Dashboard

Monitoring
  • Living Bid view (annotated BGs)
  • Pool watch (staffing, premium)
  • Trade activity log
  • ROI tracking & monthly progress
🔔

Notifications

Push, all month
  • Trade opportunities
  • Premium pay alerts
  • No-fly list conflicts
  • Family calendar updates
👨‍👩‍👧

SO View

Partner / Family
  • When pilot is home/away (no jargon)
  • Add date constraints
  • Schedule change notifications
05

Progression

What success looks like over time.

Month 1
  • First Copilot Briefing reveals 2-3 blind spots
  • 'Holy shit' moment
  • Voice walkthrough for bid entry
  • Shares ROI card
Month 2
  • Check-in: 3 minutes
  • 'Same as last month, protect the 14th'
  • 1 new optimization found
  • Memory starts confirming patterns
Month 3
  • TOGA proactively suggests standing bid changes
  • 'You always trade away Fridays — add to standing bid?'
  • Entry: 20 min of targeted edits
  • No walkthrough needed
Month 6
  • Seasonal awareness kicks in
  • 'Summer months: you prefer max days off'
  • Check-in: 2 minutes
  • Collective intelligence improving
Month 12
  • Full annual cycle understood
  • Predicts holiday bidding strategy
  • Surfaces new options from seniority gains
  • Pilot barely thinks about scheduling
06

Data Architecture

What feeds the engine — inputs, storage, and learning.

DATA SOURCES

Pilot-Provided (No AUP Risk)
Pilot profileOnboarding conversation
Seniority numbersPilot provides
LPA/LCF/LCC paramsPDF upload
Bid award reportPDF upload
Standing bid groupsPilot describes
Family calendarCalDAV sync
Monthly preferencesMonthly check-in
Critical Gap
Trip pool (pairings)See access paths below
Future (Post-Authorization)
Pool display dataCCS / TripTrade+
Trade executionCCS / TripTrade+
Real-time scheduleCCS / TripTrade+

TRIP POOL ACCESS PATHS

MVP: Pilot-providedNone RISK

Screenshots or bookmarklet

Scale: Per-pilot
ALPA partnershipNone RISK

Union committee shares data

Scale: All categories
Jeppesen/Boeing APINone RISK

Commercial data partnership

Scale: All airlines
Chrome extensionMedium RISK

Reads PBS DOM

Scale: Per-pilot

LEARNING MEMORY

Month 1
Stated preferences
Month 2
Confirmed patterns
Month 3
Discovered patterns
Month 6
Seasonal awareness
Month 12
Full annual cycle
Month 24
Career trajectory