REMAPPINGTRAVELWITHAGENTICAI

TableofContents

1Introduction4

2AI’sgrowingbutnotfullyrealizedroleintravel6

3Agentic:ThenextchapterinAI’sevolution11

4Theagentic-AI-poweredfutureofcustomerexperienceintravel15

5HowagenticAIcanimprovetravel’sinternalworkflows20

6LaunchingandacceleratingagenticAIadoption25

AboutSkift

Skiftisthelargestindustryintelligenceplatformpro-vidingmedia,insights,andmarketingtokeysectorsoftravel.SkiftdeciphersanddefinestrendsforglobalCEOsandCMOsacrossthetravelindustrythroughacombinationofnews,research,conferences,andmarketingservices.

AboutMcKinsey&Company

McKinseyisaglobalmanagementconsultingfirmcommittedtohelpingorganizationsacceleratesus-tainableandinclusivegrowth.Weworkwithclientsacrosstheprivate,public,andsocialsectorstosolvecomplexproblemsandcreatepositivechangeforalltheirstakeholders.Wecombineboldstrategiesandtransformativetechnologiestohelporganizationsinnovatemoresustainably,achievelastinggainsinperformance,andbuildworkforcesthatwillthriveforthisgenerationandthenext.

Theworkisindependent,reflectstheviewsoftheauthors,andhasnotbeencommissionedbyanybusiness,government,orotherinstitution.

Acknowledgements

ThereportdrawsonjointresearchcarriedoutbetweenMcKinsey&CompanyandSkiftResearch,includingexecutiveinterviewsacrossthetravelandtechnologysectors.

Theauthorswishtothankanumberoftravelandtechnologyexecutiveswhogenerouslysharedtheirperspectives,includingMarilynMarkham(VicePresidentofAI&AutomationStrategy,AmericanExpressGlobalBusinessTravel);NaveenManga(CTO,Marriott);RodrigoAcunaAgost(HeadofResearch&AI,Ama-deus);MatthiasKeller(CPO,Kayak);AtyabBhatti(CEOandCofounder,SkyLink);KurienJacob(PartnerandManagingDirec-tor,HighgateTechnologyVentures);GiladBerenstein(Founder,BrookBayCapital,LLC);CaraWhitehill(VicePresident,Thayer);DrewPinto(ExecutiveVicePresident,Marriott);ParveenChan-derKumar(CEO,IHCL);AlexMcGowan(Chiefoperationsandservicedeliveryofficer,CathayPacific);BrenoHelfsteinMoura(Headofhospitality,Palantir);DavidNeeleman(CEO,Breeze);ShilpaRanganathan(CPO,Expedia);andRobRansom(CSO,BookingHoldings).

ThereportteamalsobenefitedfromthesupportofMcKinseycolleaguesSalArora,JustinJ.Zhang,JosephIm,KevinMajor,WilliamHawley,AnkitMital,AnkurPuri,KaitlinNoe,DarrenRivas,SarahSahel,KamilaCieslak,AlessandraPowell,andNadyaSne-zhkova,allofwhomplayedaninstrumentalroleincreatingthisreport.TheauthorswouldadditionallyliketothankSkift’sTaylorSlatteryandMcKinsey’sSethStevensonforeditorial,externalrelations,andcommunicationssupport.

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY3

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY4

Introduction

A

genticAIcouldupendthetravelindustry.TravelandhospitalityorganizationsshouldexplorehowagenticcanhelpcatalyzeAI’stransformativepotential.

genAI’ssignificanceformostcorporationsremainsunde-termined:McKinseyresearchindicatesthat78percentofcompaniesreportusinggenAI,yetmorethan80percentofcompanieshaveseennomaterialcontributiontoearningsfromtheirgenAIinitiatives.Thissuggeststhattherecontinuestobegreatenthusiasmaboutwhatthetechnologycandoandgreatuncertaintyabouthowbesttoextractvaluefromit.

AgenticAIisthebuzziestnewtechtoolonthescene,anditoffersthrillingcapabilities.WhilegenAImostlyfunctionsasanadviserbyprovidingwell-informedcounsel,agenticAIcanfunctionmoreasadirectreportbyaccomplishingtasks.Ithastheagencytomakedecisionsandthenautonomouslyactonthem.Itcanidentifyproblems,findfixes,andapplysolutionsallonitsown.It’ssmartandtireless,aself-starterandgo-getter,requiringonlylimitedhumanoversight.Per-hapsmostimportant:AgenticAIcanserveasaninterfacethatcouldhelpcompaniesharnessthefullpowerofAI.

Thetravelindustryhasbeennostrangertotechupheaval.Tectonicshiftsoverthepastseveraldecadeshavechangedthewaysthatweplan,book,andexperienceourjourneys—whilealsodisruptingthecompaniesthathelpbringthosejourneystolife.

Whilesometechinnovationscreateclearandcapturableopportunitiesfortravelindustrygrowth,othersoffertanta-lizingpotentialthatbumpsupagainstreal-worldchallenges.Withthearrivalofeachsuccessivetechtrend,industrystake-holdersareforcedtoasksomequestions:WillthisflashyinventionbringdelighttotravelersanddeliverROItotravelcompanies?Orwillitburntime,money,andeffortwhilegeneratinglittleinthewayofconcreteresults?

ConsidergenAI,aheraldedtechnologythathasgarneredconsiderableattentionfrominvestors,executives,andboard-rooms.WithintwomonthsofChatGPTusheringgenAIintothepublicconsciousnessin2022,thesoftwarehadalreadyattractedmorethan100millionusers.Butafewyearslater,

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY5

1

2

3

4

5

6

》INTRODUCTION

AgenticsystemscouldholdthekeysthatatlastunlocktheAI-fu-eledproductivityandefficiencygainsthathavesofarproved

elusiveformanyorganizations.AgenticAImakespossibletrue

processreinvention,notjusttaskautomation,byembeddingagentsacrosscomplexverticalworkflowsandfunction-specificusecases.Itcanboostspeed,personalization,andresiliencewhileopeningnewrevenuestreams.

GiventhetremendouspotentialofagenticAI,travelandhospital-itycompaniesarebeginningtoexperimentwithit.Buttorealizethetechnology’sfullimpact,organizationswillneedtocreatenewAIstrategies,governance,andinfrastructure—alteringcorebusinessprocessesandwaysofworking.Companieswillhavetoshiftfromscatteredpilotstoenterprise-scaletransformationsarchitectedbycross-functionalteamsandchampionedbydeeplyengagedC-suiteleaders.

WhatdoesthetravelindustryneedtoknowasitpondersgoingallinonagenticAI?Howsuitedisthistechnologytotheuniquedynamicsofthetravelsector?Whichconsumer-andnon-con-sumer-facingusecasescouldoffertravelandhospitalityplayersthemostROIintheneartermandoverthelonghaul?HowcanorganizationsintegrateagenticAIinwaysthatwillallowittodelivermaximumbenefittocompanies,workers,andconsumers?

Thisreport,acollaborationbetweenMcKinseyandSkift,exam-

inesthepotentialforagenticAItounleashthefullvalueofAIin

travel.Basedonanalysisincorporatingsurveysof1,002travelers

and86travelexecutives,thereporttracesthedigitalprologuethat

broughttheindustrytothismoment,detailsagenticAI’strans-

formativecapabilitiesforbothtravelersandtravelcompanies,

andprovidesablueprintfortravelandhospitalityleaderswho

areeagertoacceleratetheirorganizations’agenticAIinitiatives.《

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY6

AI’sgrowingbutnotfullyrealizedroleintravel

W

avesofnewtechnologyhaveperiodicallywashedoverthetravellandscape,leavingitforeveraltered.Inthe1960sand’70s,globaldistributionsystems

madeitpossiblefortravelagentstoaccessreal-timeairlineinventoryandfares,therebyenablingglobalscalabilityandboostingthegrowthofcorporatetravelmanagement.Inthe1990s,onlinetravelagenciesextendedthatsamepowerdirectlytoconsumers,forcingmanytravelagentstoreframetheirroles.Thewidespreadadoptionofsmartphonesinthe2000sledtomoreon-the-godecision-makingfortravelerswhilealsocreatingnewopportunitiesfortravelcompaniestogatheruserdataandpushouttargetedoffers.Otherongoingdevelopments,rangingfromautonomoustaxistoroomservicerobotstovariousforms

ofAI,continuetoreshapetheindustry.

Inrecentyears,AIhasopenednewhorizons.Thetechnologybuildsonmoderndigitaladvances,manyofwhichhavehaddirecteffectsonthetravelsector.Deterministicprogramminglanguagelaidthegroundworkforself-serviceairlinekiosksthatautomatethecheck-inprocess.Diagnosticandpredictiveanalyticshaveallowedhotelstotrackroomturnoverratesandairlinestoanticipateweatherdelays.Machinelearningenableddynamicpricingofticketsandstays.ThesebuildingblocksformthefoundationthatAIisconstructedon.

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY7

1

2

3

4

5

6

》AI’SGROWINGBUTNOTFULLYREALIZEDROLEINTRAVEL

Exhibit1

Advancesindigitaltechnologyhavetransformedthetravelindustryinthelast~75years

Newestwaveoftheevolution

2010-2015

2010-2020

2000-2010

AgenticAl

2024-pres.

GenerativeAl

2022-pres.

Machinelearning

2012-2020

1.IBMandAmericanAirlines|2.SchipolAirportandKLM|3.E.g.,StarwoodHotels|4.E.g.,Tomorrow.ioandJetBlue|5.E.g.,MedalliaandHilton|6.E.g.,Expedia

1

3

4

5

6

2

》AI’SGROWINGBUTNOTFULLYREALIZEDROLEINTRAVEL

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY8

AI’sincreasingtravelindustrypresence

AIhasseenacceleratingadoptioninthetravelindustry.Onlyabout4percentofcompaniesintheSkiftTravel200(asubsetofthelargestpubliclytradedtravelcompaniesintheworld)mentionedsomeformofAIintheir2022annualreports.By2024,thathadrisento35percent.

Meanwhile,AI-basedtravelstart-upsareattractingasurgeofventurecapitalinterest.In2023,accordingtoSkifttracking,onlyabout10percentoftravel-industry-relatedventurecapitalfund-ingwenttotravelstart-upsthatprovidedAI-enabledofferings.Bythefirsthalfof2025,thatfigurewas45percent.“Atleast70percentofourportfolioisusingAI,”saysKurienJacob,manag-ingdirectoratHighgateTechnologyVentures,atravel-focusedinvestmentoutfit.“It’sgettingusedinoperations,inengineering,inproduct,andnowit’sbeingbroughttocustomers.”

Sometravelandhospitalitycompaniesseemexcitedabout theopportunitiesthatAIispresenting.Inasurveyof86mostlyUS-basedtravelexecutivesconductedforthisreport,26percent ofrespondentssaythatintroducingAItotheirorganizations hasledtocostreductionsinoperations,30percentsaythatit’s aidingfasterdecision-making,33percentsaythatit’simprov- ingcustomerpersonalization,36percentsaythatit’senabling higher-qualityoutputs,and59percentsaythatit’sincreasing employeeproductivity.Amajorityofrespondentsindicatethat adoptingAIintheirorganizationshasresulted,overthepastthree years,inmorethan6percentannualrevenuegrowthandmore than6percentannualcostsavings.(It’simportanttodistinguish thebroaderdefinitionoftheterm“AI”andthecapabilitiesthatit encompassesfromthenarrowerdefinitionimpliedbytheterm“genAI.”)

“Insomeways,it’simpossibletotracktheabsolutebottom-line impact,butthere’snodoubtweareseeingreallypositivecus- tomerandbusinessoutcomesfromAI,”saysAlexMcGowan,chief operationsandservicedeliveryofficerforCathayPacificAirways.“Forexample,wehadarecenttyphoonthatshutdownflightsfrom 5:00a.m.until6:00p.m.,withmorethan200flightsstopped duringthatperiod.UsingAIandaugmented-decision-making

tools,wewereabletorebuildourscheduleinroughlytwohours,whenhistorically—usingwhiteboardsandspreadsheets—itwouldhaveprobablytakeneighttotenhoursandbeenmuchlessprecise.”

TravelconsumersarealsobecomingmorecomfortablewithAIandareincreasinglyexposedtoAItoolsthatcanenhancetravelexperiences.IntheJune2025SkiftUSTravelTrackerSurvey(whichcollectedresponsesfrommorethan1,000UStravelers),morethanhalfofrespondentsreporthavingusedAI-basedplanningtoolsfortravel.Andinjustthepastyear,useofChatGPTfortravelplanninghasincreasedconsiderably.

1

3

4

5

6

2

》AI’SGROWINGBUTNOTFULLYREALIZEDROLEINTRAVEL

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY9

Exhibit2

HowfamiliarareyouwithAl-basedtravelplanningtools(e.g,chatGPT,GuideGeek,TripGen)?veryfamiliar-IUsethemregularlyfortravelplanning

36%

Familiar-Ihaveusedthemoccasionally

24%

somewhatfamiliar-I'veheardofthembuthaven'tusedthem

Notfamiliar-Idon'tknowmuchaboutthesetools

15%

Neverheardofthembeforethissurvey

source:USTravelTrackersurvey,conductedbetweenJune3,2025,andJuly5,2025,(N=1002)

Exhibit3

HaveyouusedchatGPTorasimilarAl-basedapptoassistinplanninganyaspectofyourtrip?

20242025

30%

25%

25%

22%

21%

21%

13%

13%

yes,extensivelyyes,occasionally

source:USTravelTrackersurvey,conductedbetweenJune3,2025,andJuly5,2025,(N=1002)

1

3

4

5

6

2

》AI’SGROWINGBUTNOTFULLYREALIZEDROLEINTRAVEL

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY10

Criticalinformationgetslockedinsidedisparatelegacyinfra-structures,thwartinginteroperability.Thelackofcentralizeddataownershipacrossthetravelecosystemlimitsthenet-workeffectsandfeedbackloopsthattypicallyaccelerateAIperformance.Asaresult,itcanbeexceptionallychallengingfortravelcompaniestotraineffectiveAImodelsortodeliverpersonalized,real-time,AI-poweredexperiencesatscale.

.Travelcompaniestendtofavorinvestmentinhumanintercon-nectioninsteadoftechinnovation.Atthe2025SkiftData+AISummit,ChrisSilcock,presidentofglobalbrandsandcom-mercialservicesforHilton,said,“It’snotanunfairsuggestionthathotelshavebeenbehindthecurveintermsofAIadoption.It’sbecausewehavebeensofocusedontheexperience—thehuman-to-humaninteraction.”

Someofthetravelsector’swarinessregardingAImightbeattributedtotheindustry’sgeneralviewthatitspecializesinservice,nottechnology.Newtechcapabilitiesareoftenseenasenablersbutnotascorebusinesscomponents.Asaresult,techtalentandtechinvestmentcanlagbehind.

Thegoodnewsisthatanewwaveoftechadvancementcouldaddressbothconcerns.Itcouldhelpthesectortoovercomemanydatahurdlesandstrengthensomeofthehumanele-mentsoftravel—whilealsounlockingimportantusecasesandefficiencyimprovements.

ChallengesthathinderAI’sdeploymentintravel

Despiteallthisburgeoningenthusiasmandadoption,however,thetravelandhospitalitysectorsstillappeartolagbehindothersintermsofAImaturity.Thisfinding,suggestedbydiscussionswithindustryleaders,issupportedbyvariousdatapoints.Forinstance,ina2025surveyfromEpsilonDataManagement,11percentoftravelleaderssaythattheirorganizationsdon’tuseAIatall—almostdoubletherateofnonusethatthesurveyobservesinotherindustries.Employmentfiguresalsoindicatethatthetravelindustryhasn’trampedupitstechhiring:USCensusBureaudatarevealsthattravelsuppliers,suchasairlinesandaccommo-dationcompanies,employonly1percentoftheirworkforcesintechroles,andcompaniesinotherindustriestypicallyemploycloserto4percentoftheirworkforcesintechroles.

Travelcompanies’AI-basedeffortshavelargelycenteredoncreatingenterprise-widecopilotsandchatbots,andsucheffortshavescaledquickly.Butforthemostpart,thesemorehorizontalinitiativeshavedelivereddiffuse,hard-to-measuregains.Verticalusecasesthatarefunctionspecifictothetravelsectorcouldbemoretransformative,butthevastmajorityofthesemorefocusedexperimentshaveremainedstuckinpilotmode.

WhyhasthetravelindustryfacedchallengesasitattemptstodivedeeperintoAI?Twopossibilitiesemergeascentralstumblingblocks:

.SiloeddataandincompatiblesystemsmakeusingAImoredifficult.Thetravelindustry(andespeciallyhospitality)ishighlyfragmented,cobbledtogetherinpartfromcountlesssmalltomedium-sizebusinessesspreadacrossnearlyeverycountry.Thisleadstosiloeddatathat’sinconsistentinitsparametersanddifficulttounify.Asingleconsumercouldinteractwithmorethan100differenttravelplatformsorservicesinjustoneyear.Likewise,onetravelcompanycouldhavedataaboutasinglecustomerlodgedwithinadozendifferentoutdateddatasystems.

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY11

Agentic:ThenextchapterinAI’sevolution

njustthetimesincethepublicationofthe

2023McK-

inseyandSkiftreportonAIintravel

,therehavebeensomeremarkableAIinnovations.Perhapschiefamongthese

istheadventofagenticAI,acutting-edgetechnologypoisedtorevolutionizethetravelindustry.AfewfeaturesdistinguishagenticAIfrompreviousformsofAI(includinggenAI):

I

.AgenticAIcanautonomouslymakedecisionsandtaketheinitiativetoaccomplishgoals,usingmultistepreasoningwhileundertakingcomplexactions.

.Toexecutetasks,agenticAIcancallonexternaltools,APIs,andsystems.

.AgenticAIcanstoreandrecalllong-term,structuredmemoriesthattrackcontext,progress,anduserpreferences,allowingittodeeplypersonalizeitsresponsesandtohandlerequestsspreadacrossmultiplesessions.

WhilegenAIhasmostlyplayedareactive,advisoryrole—answer-ingresearchquestionsandofferinghelpfulsuggestionswhenspecificallypromptedtodoso—agenticAItakesafarmoreproactiveposture.AnAIagentcancarryoutacomplicatedtaskendtoendandcanevendirectteamsofotherAIagentstoworktogetherontheproject.Agenticcanmonitorasituation,under-standonitsownwheninterventionisnecessary,andthendeveloparesponsiveplanandenactitwithlittlehumaninvolvement.Itcanalsorememberwhosomeoneisandwhatthey’veneededinthepastandthenincorporatethatunderstandingintothepresent.

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY12

1

2

3

4

5

6

》AGENTIC:THENEXTCHAPTERINAI’SEVOLUTION

Exhibit4

AgenticAlelevatesexistingAltechnologies

Non-exhaustive

LevelHighLOW

GenAl

primaryfunction

Decision-makingprocess

Adaptability

Memory&Learning

Multi-systemsintegration

Humaninvolvement

●Rulessetbyhumans

scusecasesexamples

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY13

1

2

4

5

3

6

AgenticAI’scapabilitiescouldenableexcitingnewpossibilitiesinalmostanysector.ButwhyisagenticAIespeciallysuitedtodisruptingthetravelindustry?

.Unliketraditionalautomationtoolsthatcanonlyaccessstruc- turedAPIs,agenticAIcannavigateuserinterfacesdirectly, clickingthroughwebsitesandlegacysystemsmuchasa humanwould.Thisenablestheintegrationofinformation infragmentedsituationsthatprevioustechnologiessimply couldn’thandle.Byactingasconnectivetissueinthisway, agenticAIcanretrieveorupdatedataacrosssilos,performmultistepworkflows,and,importantly,savehumansfrom doingtediousinterpretationsbetweendisconnectedsystems.“Intravel,youhaveallthesesystemsthatdon’ttalktoeach other,”saysMarilynMarkham,vicepresidentofengineering andAIstrategyatAmericanExpressGlobalBusinessTravel (AmexGBT).“Humanscansometimesfeellikethey’rehaving tospendtimeactingas‘middleware’thatcallonthevarious applicationsandtransferresultsfromoneovertothenext.Wewanttofreeourteamstofocusoncomplexcustomerqueries,andonmorevaluabletasks.”

.AgenticAItendstoshineinsituationsthatdependonretriev-ingandreasoningacrosstwodifferenttypesofdatasources:structureddatasources(suchasflightschedules,pricingtiers,andbookingrecords)thatarevitalforexecutingactionsandunstructureddatasources(suchassupporttranscripts,customerreviews,andtravelblogs)thatarecrucialforunder-standingnuancedcontext,preferences,andsentiment.Thisisafrequentlyseenblendofcomponentsinthetravelindustry,andagenticAIcaneffortlesslycombinethesedisparateele-mentsbecauseitcanaccessbothtypesofdatadynamically,interprettheminrealtime,andactaccordingly.Thiscanallowend-to-endpersonalization,decision-making,andautomationatscale.

》AGENTIC:THENEXTCHAPTERINAI’SEVOLUTION

.Theidealagenticusecasehasaclearobjectivethatcanonlybeaccomplishedbyusingopen-ended,multistepreasoninganddecision-making.Manytravelindustryusecasesfallintothiscategory.Fewobjectives,forexample,areclearerthanreturningastrandedtravelerbackhome.Tomanageadisruptedtravelitinerary,however—repletewithacanceledflight,unexpecteddelay,andsuddenneedforaccommoda-tion—canrequiremultistepcoordinationacrossavarietyofchannels.AgenticAI,initsfullyrealizedform,couldrebooktheflight,extendthehotelreservation,cancelthecarpickup,andupdateallbilling.

.AgenticAIthriveswhenpresentedwithahighvolumeoftasksthataregenerallyrepetitiveinbasicstructurebutstillrequirepersonalizationbasedonindividualcontextoruser-specificdata.Manyworkflowsintravelandhospitality—forinstance,bookingLondon-to-NewYorktripsforpartnersofalawfirmwithheadquarterslocatedinbothcities—followacon-sistentpatternanddon’tneedtobecompletelyreinventedeachtime,yettheyalsocan’tbesufficientlyhandledusingaone-size-fits-allapproach.

REMAPPINGTRAVELWITHAGENTICAISKIFT+MCKINSEY&COMPANY14

1

2

3

4

5

6

》AGENTIC:THENEXTCHAPTERINAI’SEVOLUTION

ConsiderthehypotheticalexampleofSally,aLondonpartneratagloballawfirmwhoneedstoflytoNewYorktomeetwithcol-leagues.Hundredsofpartnershavemadethissametripbeforeher,allfollowingthesameroutebookedagainstthesamesetofcompanytravelpolicies.Sally,however,hasafewspecializedrequirements:Shecan’twalklongdistancesandneedsassis-tanceattheairporttogettohergate,shepreferssittinginanaisleairplaneseatwhereshehasmoreelbowroomtowork,shealwaysusesthesameairlineforflightstotheUnitedStatestogainloyaltypoints,shelikeshotelroomswithtwoqueenbedsformorespace,andsheavoidshotelroomsontopfloorsduetoherfearofheights.

Agenticsystemsarewellequippedtodeliveratripsuchasthis,basingtheiractionsoninformationderivedfromcustomerinter-actionsthatrequireadaptive,high-touchengagement.AgenticAIcanconnectfragmentedsystems,combinestructuredandunstructureddata,tweakarepetitivetaskbasedoncontext,andnavigatethetravelindustrymazetoachieveallofSally’sobjectives.Whenfullydevelopedanddeployed,agenticAIcouldcreateapersonalizeditineraryforSallythattailorsthetriptopref-erencesshementionsinachatsession,incorporatespreferencesrememberedfrompreviouschatsorgleanedfromotherdatasources,reflectsupdatedinformationandfeedbackfrommorerecenttripstakenbyotherlawfirmpartners,andalignswithherfirm’slatestpolicies.AgenticcouldthenseamlesslycoordinateacrossmultiplebookingsystemstoensurethatSally’sentiretripcouldbereservedatthepressofabutton.

Butconsumer-facingusecasesaren’ttheonlytravelindustrysituationsinwhichagenticAIcouldthrive.Itcouldalsobeappliedtointernalworkflows.ConsiderthehypotheticalexampleofMiles,ahousekeepingmanageratalargehotelwhojugglesstaffsched-uling,roomreadiness,andunexpectedguestrequests.HeadingintoaweekendwithalargeSaturdaynightweddingplanned,Milesisconcernedaboutkeepingthingsrunningwithoutahitch.

IfahotelhasimplementedanagenticAIsystem,thatsystemcouldautonomouslyanticipateaspikeinlate-checkoutrequestsbasedonbothstructureddata(suchaspatternsofguestsinpreviousreservedweddingblockswho’verequestedlatecheckouts)andunstructureddata(suchasincomingfeedbackfromcurrentguests).AnAIagentcouldsuggestoptimizedstaffingshiftsandcoverageforMilestoapprove,setupcontingencyplansforlatecheckouts,andestablishcoordinationwiththehotel’sbookingsystem,roomstatusdatabase,andpayrollsoftware.Onthedayofthewedding,asguestrequestsrollin,theAIagentcould