The Deloitte Prototypes.

The Borrowed Fleet.

LA28 must borrow a second bus fleet from the rest of America, run it for a month, then give it all back.

The plan of record for LA28 is the boldest operational promise in modern Games history: a transit-first Olympics in the most car-bound major city on Earth. Spectators reach venues without private cars;²⁰ LA Metro and the organizers estimate roughly 2,700 additional buses, nearly doubling the existing fleet, borrowed from transit agencies across the country, supported by mobility hubs, dedicated lanes, and thousands of temporary operators.¹⁹ Deloitte is not a bystander to this: it is LA28’s Official Professional Services Provider¹² and the Games’ technology integrator through 2032.³

The Borrowed Fleet makes the audacity of that plan visible, then playable, in three acts. Act one, the assembly: which agencies lend what, and at what deadhead cost; a map of America’s buses flowing west. Act two, the month: depot siting and daily route assignment across 49 venues in 18 zones, with demand surging and collapsing on the rhythm of the competition schedule.¹⁵ Act three, the return. The user holds the levers: fleet size, depot placement, how aggressive the dedicated lanes get, and watches service levels respond. Then the optimizer takes a turn and shows the assignment that covers the same demand with meaningfully fewer vehicles.

The point the prototype makes without saying it: a transit-first Games is not a slogan, it is a vehicle-scheduling problem of national scale, and whether it succeeds is decided by exactly the kind of mathematics on display. The browser holds a simplified region, solved exactly. The real instance, a national lending network crossed with a month of venue schedules and depot constraints, is precisely the workload the Strangeworks platform exists to run. One model definition, two scales: the same pattern as everything else in this series.

And the honest coda, which connects to The Long Run: when the month ends, the buses go home. What stays is what the plan chose to make permanent: the hubs, the lanes, the operating playbooks. The prototype keeps that ledger in view, because the legacy question is the real one.

How it works

• Data: reconstructed from public Metro and Games-mobility planning coverage, cited inline; clearly labeled as our model, not the official plan.
• Model: fleet assembly as an assignment problem; the month as vehicle scheduling over depot, route, and demand constraints; exact for browser-sized regions.
• The levers: fleet size, depot siting, lane priority; service-level curves respond live, with the optimizer’s answer rendered against yours.
• Scale: the full national instance runs on the Strangeworks platform; the browser and the platform share one model definition.

Why it’s unmistakably Deloitte

The firm advising LA28’s operations should be the one that makes its boldest operational promise legible to the public. It joins the Future of Mobility practice,¹⁸ the LA28 mandate, and the technology-integration role in a single artifact, two years before the world arrives to grade the result.

Audience

LA civic and transit leaders, mobility press, operations-minded fans

Form

Three-act fleet simulator with an optimal-assignment reveal

Engine

Assignment + vehicle scheduling; full national instance on Strangeworks

Moment

Every LA28 transport milestone between now and July 2028

Artifact

The fleet ledger: a map card of borrowed buses flowing west, then home

Gate to Gate.

A car-free Games day is a routing problem. Pick your sessions; the solver makes every gate, or proves it cannot be done.

When Deloitte signed on with LA28, the organizing committee’s commercial chief said the partnership would examine “the fan journey like never before.”¹² Here is the fan journey, literally: gymnastics downtown in the morning, swimming in Inglewood in the afternoon, athletics at the Coliseum at night, forty-nine venues across eighteen zones,¹⁵ and an official answer to “how do I get there” that is, by design, transit.²⁰ Whether any given day of Olympic ambition is physically possible is, today, a question nobody can answer.

Gate to Gate answers it with a clock and a graph. Choose your sessions; a time-window router over the transit network returns the day plan: lines, transfers, walking legs, and the buffer minutes between arrival and gate. Or it returns the other verdict, the honest one: infeasible, with the reason stated in one sentence: the gap between the session ending and the next gate closing is thirty-eight minutes, and the trip needs sixty-one. The summary view is the feasibility frontier: for any pair of sessions, what is combinable and what is fantasy.

For Deloitte the demo argues a thesis the Fan Data Platform implies but cannot show: personalization is not content, it is logistics. The platform knows what a fan loves;⁶ Gate to Gate knows what a fan can physically do; the actual fan experience is the intersection of the two. It ships now as a schedule-and-map model, since the session schedule is published,¹⁵ upgrades to live headways as the Games approach, and re-skins to any multi-venue event afterward.

How it works

• Model: time-window routing over a transit graph built from published schedule data; every itinerary carries explicit buffers.
• The verdict: feasible plans drawn on a day timeline; infeasible combinations explained in plain language, with the binding constraint named.
• The frontier: a session-by-session matrix of what can be combined: the shareable, argued-about artifact.
• Path to live: static schedules now, real-time headways later; the model does not change, only the data feed.

Why it’s unmistakably Deloitte

It converts the transit-first promise from a press line into a usable tool, and it is the most concrete possible expression of the fan-journey work Deloitte signed up for in 2021. Nothing else in this volume would be used by more actual ticket-holders.

Audience

Every future LA28 ticket-holder; transit and tourism press

Form

Car-free day planner with feasibility verdicts

Engine

Time-window routing on a transit graph; shares the mobility pair’s solver core

Moment

Ships on the published schedule; grows through every ticket window to 2028

Artifact

“Your Games day” transit card, plus the feasibility frontier

The Engine Room.

Run the technology of a Games day. If you do it perfectly, nobody will ever know you were there.

In February, the most important room at the Winter Olympics had no podium and no cameras. The Technology Operations Centre in Milan, designed and run by Deloitte with the organizing committee, kept accreditation, results distribution, venue systems, cloud, and cyber defense alive across two mountain ranges and a city, with remote operations centres extending it across borders.^4,5 The IOC President came by to thank the room personally. Then the Games ended, and the best technology story of the fortnight evaporated, because its whole achievement was that nothing happened.

The Engine Room makes that job playable. It’s a single-sitting simulation of one Games day in the operations chair: incidents stream in (the results feed at the oval is lagging, an accreditation scanner is down at a village gate, a credential-stuffing spike is hitting the fan login, a venue’s timing backup just went amber), each with a severity, a clock, and a dependency chain. You hold a finite bench of specialists. You decide who goes where, what gets fixed first, what waits, and what you accept will degrade quietly. The screen is deliberately calm: a wall of green tiles you are trying to keep green.

When the shift ends, a real dispatch solver replays your day and shows the optimal allocation: what the best possible operator would have done with the same bench, and the uptime gap between you and it. The reveal carries the concept’s one-line lesson: at the Olympics, perfection is the baseline, and perfection is a resource-allocation problem. Most players will lose a venue for eleven minutes and develop a permanent respect for the people who don’t.

The re-skin path is the partnership’s own roadmap: a Dakar edition this fall, then the heavyweight version: LA28’s 49 venues in 18 zones, plus Oklahoma City,¹⁵ where the prototype doubles as a recruiting artifact for the small army of technologists the home Games will need.

How it works

• Scenario engine: authored incident decks (no live data needed) with severity, decay curves, and dependencies, built from the publicly described TOC scope: results, accreditation, venue IT, cyber, cloud.
• The model: specialists × incidents as a min-cost assignment problem over time: small, exact, solved client-side in milliseconds; the same solver shape as the lab’s existing prototypes.
• The reveal: your timeline vs. the optimal timeline, side by side, with the three decisions that cost you the most.
• Artifact: a “shift report” card (uptime percentage, incidents cleared, your costliest call), built for comparing scores.

Why it’s unmistakably Deloitte

The technology-integration mandate is the largest, newest thing Deloitte does in sport, and the hardest to show. This shows it, honestly, as an allocation discipline rather than a magic trick, and it banks the Milano Cortina story while it’s still warm, with LA28 as the sequel.

Audience

Fans, tech press, future Games technologists

Form

Single-sitting operations simulation with a solver reveal

Engine

Time-indexed assignment/dispatch model; reuses the lab’s solver core

Moment

Milano Cortina retrospective now; Dakar edition in October; LA28 flagship

Artifact

“Shift report” score card

First Sport.

Three hundred and fifty events. One of them was made for you. Five questions and you’ll meet it.

The Fan Data Platform exists because of a structural truth: the Olympics is the only sporting event where almost every viewer is a newcomer to almost every sport. LA28 will stage 36 sports and 51 disciplines across more than 350 medal events, including cricket, flag football, squash, and lacrosse, which most American viewers have never watched once.¹⁵ The IOC pays Deloitte to solve this at industrial scale: 32 million fan profiles, four engagement segments, generative-AI personalization.⁶ All of it invisible. From the outside, personalization just looks like luck.

First Sport is that machinery turned into a two-minute public matchmaking ritual. Five questions, none about sport: What gets you: raw speed, controlled grace, brute power, deep strategy, or honest chaos? Alone or as a team? Do you want a verdict in seconds or a story over hours? Underdogs or dynasties? Watching, or secretly wishing it were you? From those, a transparent matching model walks the full Olympic program and returns your ranked matches, not just the sport, but why: “You said chaos, short verdicts, underdogs: you are a ski-cross person. Here’s what to feel when the gate drops.”

Each match comes with a starter kit: the three rules that matter, one thing the commentators won’t explain, and, borrowing the partnership’s own campaign language, a first to watch for in that sport’s next edition.¹⁶ The output card is “your opening ceremony”: your five sports, your sessions, the schedule that makes you a fan of something new. The pilot follows the platform’s own precedent: the real CDP was trialed at a Youth Olympic Games before Paris; First Sport pilots at Dakar 2026, the first Olympic event ever held on African soil, then scales to the LA28 audience.^6,17

How it works

• Model: a transparent trait-vector match between five answers and ~50 disciplines, each hand-scored on the same axes; no account, nothing stored; the model is the demo, not the data.
• Content: a starter-kit card per discipline (rules, the thing to feel, the “first” to watch), authored once, reused every Games.
• The dial: answers are editable after the reveal, so users can watch their match shift: personalization made legible.
• Artifact: “your opening ceremony” card: five sports, sessions, and the one-line why for each.

Why it’s unmistakably Deloitte

It is the Fan Data Platform’s thesis (segmentation, then personalization, then a bigger audience) performed in public at toy scale, with the logic exposed. Every new fan it converts is literally the metric Deloitte’s platform is paid to move.

Audience

Every casual viewer; the LA28 newcomer wave

Form

Two-minute matchmaking interactive

Engine

Transparent trait-vector matching, fully client-side

Moment

Pilot at Dakar 2026 (Oct 31–Nov 13); flagship for LA28

Artifact

“Your opening ceremony” schedule card

The Eleventh Event.

A decathlon has ten events. The eleventh is the career after, and it’s scored, too.

This is the partnership’s most human work-stream, and Deloitte’s fingerprints on it are specific: the firm interviewed hundreds of athletes for the IOC, wrote the Athlete Employability Framework and Handbook, and helped launch Athlete365 Mentoring, where eleven of Deloitte’s own retired Olympians and Paralympians mentor athletes facing the cliff every athlete faces.⁸ One of Deloitte’s consultants on that engagement put the problem in a sentence: “I’ve seen people that win gold medals. And then they sit down on the couch for a year.” The Australian data behind Deloitte’s Ignite program is blunter still: nearly half of Brisbane 2032 hopefuls have considered quitting sport over money and mental-health pressure.¹⁴

The Eleventh Event turns the framework from a PDF into an instrument, using the most Olympic scoring metaphor there is: the decathlon points table, which converts wildly different performances into one honest currency. An athlete, or anyone whose résumé doesn’t look like a résumé, enters their sport, role, level, and years. The model converts that history into scores on ten transferable capabilities: performing under public pressure, deliberate-practice discipline, operating inside team systems, coachability, data fluency from training analytics, comeback management, and so on. Every scoring table is visible; nothing is a black box.

Then it matches: which job families value exactly this profile, where the genuine gaps are, and what closes each gap (the follow-through that makes it real rather than flattering), linked to the mentoring and employability resources Deloitte already builds for the IOC. Employers get the inverted view: load a role, see which athletic profiles over-index for it, and read the one-page case for why the “non-traditional” résumé in the pile might be the best one. The share card is the capability decathlon itself, ten bars and a total score, designed for LinkedIn, where second careers actually start.

How it works

• Model: capability scoring tables (sport × role × level × years) published in full on the page, decathlon-style; then a weighted match against ~20 job families.
• Honesty rule: scores come with floors and ceilings explained: the tool never claims sport alone closes a skills gap; gaps are the point.
• Two doors: athlete view (profile → matches → gap plan) and employer view (role → profiles → the case for the athlete résumé).
• Artifact: the capability-decathlon card: ten scored bars, one total, built for LinkedIn.

Why it’s unmistakably Deloitte

It operationalizes a framework Deloitte literally wrote, in the voice of a firm that employs more than forty Games competitors and runs athlete-hiring programs on two continents.^9,14 It serves athletes, employers, and Deloitte’s own recruiting brand in one move, and it keeps working every year between Games, which no medal-pegged activation can say.

Audience

Athletes in transition, employers, early-career talent

Form

Skills-translation instrument with published scoring tables

Engine

Scoring + weighted matching; the lab’s selection-solver DNA

Moment

Evergreen; re-pegs at every Team Deloitte and Ignite announcement

Artifact

Capability-decathlon card for LinkedIn

The Playbook.

A hundred and sixty AI ideas want into the stadium. Sequencing them is the strategy.

When the IOC launched the Olympic AI Agenda in April 2024, Deloitte wasn’t quoted about it; it was inside it: a seat in the 16-expert working group, an evaluation of more than 160 candidate AI use cases across the Movement, the prioritization of that portfolio, and the playbook for deploying it responsibly.^10,11 That work answered the question every large organization is currently failing to answer: not “what could AI do?” (everything, allegedly) but “in what order, under what guardrails, and what must be true first?”

The Playbook puts that question in the public’s hands. The screen is a portfolio room: a constellation of AI use cases drawn from the Agenda’s published focus areas (talent identification, training and injury prevention, judging support, accessibility, fan personalization, energy optimization, abuse monitoring), each card scored on readiness, impact, and risk, each carrying its trust prerequisites: bias audit, data provenance, human-in-the-loop, athlete consent. The user sets the Movement’s priorities with four weights (athlete welfare, fan growth, operational efficiency, competition integrity), and a dependency-aware sequencer rebuilds the multi-year roadmap live.

The mechanic teaches the doctrine. Crank fan growth to maximum and watch athlete-welfare use cases slide down the roadmap, and feel why that’s a values choice, not a technical one. Try to drag a flashy use case to year one and the trust gates physically block it: its prerequisites aren’t met, and the interface shows exactly which ones and why they exist. AI strategy is sequencing under constraints: the single most Deloitte sentence in this entire document, demonstrated instead of asserted. The artifact is your roadmap: the first ten use cases in your order, plus the honest line every strategy needs: what you chose to postpone.

How it works

• Content: ~40 public-knowledge use-case cards built from the published AI Agenda focus areas; our reconstruction, clearly labeled, not Deloitte’s confidential portfolio.
• Model: weighted prioritization + topological ordering over trust-prerequisite dependencies: a sequencing solver, solved instantly client-side.
• The gates: trust prerequisites rendered as physical constraints in the interface, with one-line explanations of each.
• Artifact: “your roadmap” card: ten use cases in order, with the postponed list printed at the bottom.

Why it’s unmistakably Deloitte

Responsible-AI-at-scale is Deloitte’s flag, and this is its Olympic case study made explorable. It demystifies an Agenda most fans have never heard of, and it doubles as the cleanest possible client demo: the same portfolio room works for a hospital system or an airline the day after it works for the Olympics.

Audience

Business and tech press, policy watchers, every Deloitte client

Form

Interactive portfolio-sequencing room

Engine

Weighted prioritization + dependency ordering; scheduling-solver DNA

Moment

AI Agenda anniversaries, IOC AI forums, every “AI at the Games” news cycle

Artifact

“Your roadmap” card with the postponed list

The Long Run.

The Games last seventeen days. The argument about what they were worth lasts twenty years. Simulate it.

Whether hosting the Olympics is worth it is the most contested claim in sport, and Deloitte holds a distinctive position in the argument: Deloitte Access Economics’ Going for Gold (July 2025) modeled Brisbane 2032’s opportunity across two decades (people, places, perception) using a general-equilibrium model and an explicitly independent posture.¹³ Meanwhile the hosting doctrine itself has flipped: LA28 plans zero new permanent venues; Brisbane builds on a venue stock that already mostly exists.¹⁵ The era of the white elephant is supposedly over. Supposedly is a testable word.

The Long Run is a hosting-legacy simulator. Pick a host archetype (compact megacity, regional spread, first-time host) and make the five choices that actually separate good Games from cautionary tales: build venues or reuse them; how much transport to bring forward; whether to fund participation programs or just the show; tourism strategy; how much of the budget is private. Then run twenty years. The output is not a number but a ledger over time: costs early, benefits late, with uncertainty bands that widen honestly the further out you look.

The control that makes the prototype matter is the counterfactual dial: the assumption every impact study lives or dies on, and the one almost none of them surface: how much of this would the city have done anyway? Drag it and watch “the Games built this” deflate into “the Games accelerated this,” and sometimes into less. A user who plays with that dial for ninety seconds permanently understands why two studies of the same Games can disagree by billions, and why a firm willing to put the dial on screen is the one whose numbers deserve quoting. Each archetype ships with a real-world epilogue: what the no-new-venues doctrine looked like in Los Angeles in 1984, what Sydney’s ledger looks like twenty-five years on.

How it works

• Model: a compact, parameterized legacy model (capex, usage, tourism, participation, health and brand effects) with published coefficients drawn from the academic literature and Going for Gold’s public framework; our reconstruction, labeled as such.
• Uncertainty as a feature: every projection renders as a band, not a line; the bands widen with horizon, on purpose.
• The dial: counterfactual adjustment exposed as the primary control, with a plain-English explainer.
• Artifact: the “legacy ledger”: your Games vs. the no-Games baseline, year by year, assumptions printed on the card.

Why it’s unmistakably Deloitte

It extends Deloitte’s most-cited piece of Olympic analysis into the interactive medium the hosting debate actually happens in, and it stakes the same ground the Vol. I economics concept staked: in a genre famous for inflation, the firm that exposes its assumptions owns the credibility. Every future host conversation (French Alps 2030 milestones, Brisbane’s mid-runway reviews, the 2036 bid race) re-pegs it.

Audience

Civic leaders, economics press, host-city skeptics and boosters alike

Form

Twenty-year legacy simulator with uncertainty bands

Engine

Parameterized scenario model, fully client-side, assumptions exposed

Moment

Host-cycle news: French Alps 2030, Brisbane reviews, 2036 bidding

Artifact

“Legacy ledger” vs. the no-Games baseline