Proposal · Vancouver Airport Authority

One briefing.
Every portfolio.
Every morning, from live systems.

A leadership briefing platform for YVR: a daily, sourced executive digest for each member of the Executive Team, and a plain-language way to ask the airport a question — built on the systems Sea Island already runs.

For discussion with the Executive Team and Board of Directors · September 2026 · Draft v1

What you saw

The executive digest — translated to Sea Island

The report you were shown is produced every morning for another company's C-suite: one page per officer, live numbers, sources attached, no analyst in the loop. Below is what the same product looks like with YVR's portfolios and YVR's published figures in it.

YVR Leadership BriefIllustrative · Tuesday morning
Operations · COO
97%Passengers through security in ≤15 minOn target
99.9%Connecting bags delivered on time to airlinesOn target
Departure punctuality vs. plan, last 24hLive from AODB
Avg taxi-out time · Digital Twin / NAV Canada feedLive
Finance · CFO
$717.7MFY2025 revenue (FY2024: $666.3M)+7.7% YoY
$227.8MAirport Improvement Fee — largest revenue line+3.6%
$43.8MExcess of revenue over operating expensesvs $35.9M
−7.1%Transborder passenger volume, 2025Watch item
Guest experience · Commercial · People · Environment · Technology …

Figures shown are from YVR's public 2025 Annual & Sustainability Report and Consolidated Financial Statements, used here for illustration only. In production every tile is pulled live from the system of record and links back to it.

Why now

YVR's own strategy already asks for this

The 2025–2027 Strategy commits YVR to "fully harnessing the power of digital transformation and data analytics", to growth and yield, and to reducing variability across the operation. A record year makes the case sharper, not softer.

26.9MPassengers in 2025 — all-time record, +2.7% YoY
365k tCargo in 2025, +7.4% — record volumes
$717.7MTotal revenue, up from $666.3M
15×Skytrax Best Airport in North America
Growth and yield

Record volume, thin margin

Revenue over operating expenses was $43.8M on $717.7M — about 6%. A $73.3M asset write-down turned 2025 into a $50.4M deficiency. Yield decisions now need to be made weekly, on current numbers, not quarterly on assembled ones.

Variability

Exceptions are the job

YVR beat targets on punctuality, security waits and baggage connections. Holding that at higher volume means leadership seeing the exception the morning it happens — winter ops, a transborder dip, a baggage subsystem trend — not in next month's pack.

Talent-first

Judgment, not retrieval

The Strategy invests in a talent-first organization. The scarcest asset on Sea Island is the attention of the people who can see around corners. Today too much of it is spent finding, waiting for and re-explaining data that already exists.

The problem

The systems are excellent. The seams between them are staffed by people.

YVR runs a strong, mature application landscape — the Digital Twin, the airport operational database, Maximo, Oracle financials, SuccessFactors, BI. Each has an owner. Almost every question a board member or executive actually asks spans two or three of them.

The tax

Finding out

Which system holds the answer, whose team owns it, whether the figure is current, and who to ask when it isn't.

The tax

Waiting

A question that takes four minutes to answer routinely takes four days, because it queues behind someone else's week — or the next committee cycle.

The tax

Manufacturing the pack

Board and committee material is assembled by hand from exports and screenshots. By the meeting, the numbers are three weeks old and nobody can drill into them.

This isn't a tooling gap. It's a time-of-flight problem — how long a question takes to become a decision. At an airport, that time is measured against a live operation.

Proof, not theory

It's running today — for a company that is not an airport

The digest you saw is produced by a working agent at a ~600-person North American distributor. It started as an engineering tool in February 2026 and became a business platform within months. The pattern is portable; the systems are not, which is what the rest of this deck is about.

15Systems connected to a single agent — orders, payments, inventory, CX, analytics, code
675+Leadership and staff questions answered in the first 5½ months
~6 / daySourced answers, every working day, without pulling an expert off task
DailyC-suite digest: CEO · CFO · COO · CCO · CMO · CTO, with charts and sources
What made it work

Four preconditions, all portable

  • Real system access, not summaries — reads the source of truth
  • One surface people already use — chat, not a new portal
  • Enough context to be specific — the answer lives in the seams
  • Bounded authority, visible work — read-only first, every action logged
What's different at YVR

Higher stakes, better data

An airport authority is regulated, safety-critical and publicly accountable, so governance comes first, not last (slide 9). But YVR also starts further ahead: the Digital Twin has already done the hard work of integrating operational systems into a single source of truth. This proposal sits on top of that, it does not compete with it.

Tailored to YVR

One page per portfolio, from the systems each already owns

PortfolioWhat the morning brief showsLikely sources (to confirm with the CIO)
President & CEOPassengers and cargo vs. plan; the day's top three exceptions across the operation; items heading to the BoardRoll-up of every section below
Operations & COODeparture punctuality, security wait times, baggage connections, taxi and turnaround times, winter/IRROPS status, airside incidentsAODB · Digital Twin · baggage SCADA platform · CATSA feed · NAV Canada data
Finance & CFORevenue by stream (AIF, terminal, landing, concessions, parking, rentals) vs. budget; opex run-rate; ERM watch-list; capital burn; legal & privacy mattersOracle E-Business Suite · PROPworks · Chrome River · ERM register
Airport Development & Asset OptimizationCapital program milestones and variance; Maximo work-order backlog and asset health; utilization of gates, stands and beltsIBM Maximo · project controls · Digital Twin
Commercial Data Integration & OptimizationConcession sales per enplaned passenger, parking and ground-transport yield, dwell time, tenant performancePROPworks · concession POS feeds · parking systems · BI
Innovation & CIOSystem availability, service-desk load, cyber posture, Digital Twin adoption, pilot pipeline from the Innovation HubITSM · monitoring · security tooling
People & CultureHeadcount and vacancies, overtime, safety incidents and near-misses, training completionSAP SuccessFactors · Kronos · H&S system
Communications, Environment & Indigenous Relations · External AffairsGHG trajectory vs. Net Zero 2030, noise complaints, media and community sentiment, Musqueam commitments, government filesEmissions model · noise management · media monitoring · CRM

System names are drawn from public reporting about YVR's application landscape and are placeholders until validated. The design does not depend on any one of them — each connected system is built once and reused by every portfolio.

Concretely

What this looks like on a Tuesday at YVR

Finance & Operations

"Transborder is down 7%. What did that do to AIF and to gate utilization in the US pier?"

Assembled from AODB passenger counts, AIF revenue in Oracle, and stand allocation in the Digital Twin — with links to each. In ninety seconds, not the next committee cycle.

Assets & Guest experience

"Which baggage subsystem generated the most work orders last winter, and did it touch connection performance?"

Maximo work orders joined to baggage-platform events and connecting-bag outcomes. The pattern was always there; nobody could see across three systems at once.

Environment & the Board

"Are we tracking to Net Zero 2030 given this quarter's airfield emissions, and what's the sentence for the Board?"

Taxi-time emissions from the Twin against the ESP glidepath, drafted as a paragraph with sources attached, ready for the Corporate Secretary to review.

The first two need no new AI capability at all — only reachable systems and a permission model. The third is a drafting task on top of the same data. All three are Phase 1 or 2.

How it works

The architecture, in one line

Systems of record

AODB, Digital Twin, Maximo, Oracle, SuccessFactors, PROPworks, BI. Unchanged. Still the truth. Nothing migrates.

Integration layer

One standard, permissioned, read-only way for an agent to query each system. Built once per system, reused by every portfolio and every future agent.

Knowledge layer

What was never in a system: targets and thresholds, why a KPI is defined the way it is, who owns which decision, what the Board was last told.

Briefs & questions

A morning digest per portfolio, and Q&A in Microsoft Teams (or wherever leadership already works) — staff simply ask @otto. Same permissions as the person asking. Every read logged.

Complements the Digital Twin

The Twin is YVR's real-time picture of the physical airport, and the best possible data source for this. The brief is the narrative layer above it: what changed, why it matters, what the Board needs to hear — across finance, people and commercial as well as operations.

The hard part isn't the AI

It's deciding what the canonical source is for each number, who owns it, and who is allowed to see it. That is an executive decision, not an IT ticket — which is why this belongs in front of the Executive Team and the Board.

Trust & governance

Built for a regulated, publicly accountable authority

Permissions inherit, never expand

The agent sees exactly what the person asking is entitled to see. No shadow access. Compensation, legal matters, security-sensitive operational data and commercial terms stay scoped to their owners.

Read-only first, then graduated authority

Phase 1 reads and reports. Nothing is written to any system of record. Any later capability — drafting a work order, updating a register — is added only with the data owner's sign-off and explicit human approval per action.

Everything is logged; every answer carries sources

Every read, every query, every generated figure is auditable after the fact by whichever officer the Board designates. No unsourced assertions in a decision context — if it can't cite, it says so.

Data stays in Canada; no passenger PII required

Executive KPIs are aggregates. The platform is deployed in a Canadian region under YVR's control, reviewed by the CIO's security team, and aligned to PIPEDA / BC PIPA and YVR's existing privacy program.

Meet Otto

Staff will know the tool as @otto — a river otter, for the estuary around Sea Island and the Twin Otters landing at YVR South. Otto doesn't fly the plane. He surfaces what changed overnight; the crew still make the decisions. The tool's formal name, and any Indigenous identity, is chosen with Musqueam in Phase 2 — not by us.

Suggested oversight: Finance and Audit Committee for data integrity and controls; Governance Committee for the access and privacy policy. Deliberately left for this room: AI use policy, model vendor selection and residency, and which systems are out of scope entirely.

The roadmap

Three phases, each with a gate. Nobody buys the whole thing today.

Phase 1 · 90-day pilot

Connect

Two portfolios — Operations and Finance — get a real morning brief from three live systems.
  • AODB, Oracle financials, one operational feed (Twin or baggage)
  • Permission model, audit log, named data owners
  • Daily brief to the COO and CFO; weekly to the CEO
  • Baselines captured: time-to-answer, pack prep hours
Phase 2 · Months 4–8

Compose

Every portfolio has a page. Cross-system questions get answered in the thread.
  • Maximo, SuccessFactors, PROPworks, emissions
  • Plain-language Q&A for the Executive Team
  • Threshold and target logic captured per KPI
  • First committee pre-read drafted from live data
  • Formal name and identity chosen with Musqueam
Phase 3 · Months 9–12

Compound

Board material is a read, not a build. Exceptions surface themselves.
  • Board and committee packs assembled from live systems
  • Standing monitors: yield drift, punctuality, asset backlog, GHG
  • Scenario questions on top of Digital Twin data
  • Extend to Sea Island partners where agreed

Each phase ends with a demonstration to the Executive Team and a go/no-go on the next, with measured results against the Phase 1 baselines.

Scope discipline & accountability

What it is not — and how we'll know it worked

Not a headcount plan

The return is speed and capacity for the people already here. Sold internally as cost-cutting, the experts whose knowledge it needs will quietly decline to contribute.

Not a rip-and-replace

Every system of record stays exactly where it is. This is connective tissue on top of the Twin, Maximo and Oracle — not a new platform to migrate to.

Not a replacement for judgment

It assembles context and drafts. Deciding what YVR should do about the context remains the job of the Executive Team and the Board.

Not a free-for-all

One integration standard, one permission model, one audit trail. Departments buying disconnected AI tools is the failure mode this prevents.

Days → minutesTime-to-answer on cross-portfolio questions, from ask to sourced answer
Pack hoursStaff hours to prepare Board and committee material, per cycle
Weekly activesExecutives and their teams actually using it, by portfolio
ZeroPermission or data-handling incidents — the non-negotiable
The ask

What we need from this room

  1. An executive sponsor and an executive owner per connected system. Not an IT contact — an accountable officer who decides what the canonical source is and who may see it. This is the single biggest determinant of whether the pilot lands in 90 days.
  2. The CIO's team as partner, not gatekeeper. Security review, hosting in a Canadian region, and read-only access to three systems for the pilot. The Innovation Hub is the natural home.
  3. Approval of a 90-day pilot. Budget is three lines — integration work, platform and model costs, one dedicated engineer — tabled as a one-pager. Phases 2 and 3 return to this group with evidence.
  4. One rule going forward: new systems are procured only if they can be connected. No more islands.

Decision requested: approve the pilot, name the sponsor and the three system owners. Everything after that comes back to you with measured results.

Twelve months from now

What we'd like the Board to be able to say

"Nobody at YVR waits a committee cycle for a number that lives in a system we already own."

"Every executive starts the day knowing what changed on Sea Island overnight, and why."

"Our Board pack was pulled from live data the morning of the meeting — and we could drill into any figure in the room."

"We did it under our own governance, in Canada, read-only, with every query on the record."

The pattern is proven. The systems are YVR's. The only open question is whether to run a 90-day pilot — and who owns it.

Vancouver Airport Authority · Leadership Briefing Platform · Proposal draft v1