ARTIFICIALINTELLIGENCE

INACTION

SELECTEDADBINITIATIVES

INASIAANDTHEPACIFIC

MAY2024

ASIANDEVELOPMENTBANK

ARTIFICIALINTELLIGENCE

INACTION

SELECTEDADBINITIATIVES

INASIAANDTHEPACIFIC

MAY2024

ASIANDEVELOPMENTBANK

CreativeCommonsAttribution3.0IGOlicense(CCBY3.0IGO)

?2024AsianDevelopmentBank

6ADBAvenue,MandaluyongCity,1550MetroManila,Philippines

Tel+63286324444;Fax+63286362444

Somerightsreserved.Publishedin2024.

ISBN978-92-9270-687-6(print);978-92-9270-688-3(PDF);978-92-9270-689-0(ebook)

PublicationStockNo.SPR240246-2

DOI:

/10.22617/SPR240246-2

TheviewsexpressedinthispublicationarethoseoftheauthorsanddonotnecessarilyreflecttheviewsandpoliciesoftheAsianDevelopmentBank(ADB)oritsBoardofGovernorsorthegovernmentstheyrepresent.

ADBdoesnotguaranteetheaccuracyofthedataincludedinthispublicationandacceptsnoresponsibilityforany

consequenceoftheiruse.ThementionofspecificcompaniesorproductsofmanufacturersdoesnotimplythattheyareendorsedorrecommendedbyADBinpreferencetoothersofasimilarnaturethatarenotmentioned.

Bymakinganydesignationoforreferencetoaparticularterritoryorgeographicareainthisdocument,ADBdoesnotintendtomakeanyjudgmentsastothelegalorotherstatusofanyterritoryorarea.

ThispublicationisavailableundertheCreativeCommonsAttribution3.0IGOlicense(CCBY3.0IGO)

/licenses/by/3.0/igo/.Byusingthecontentofthispublication

,youagreetobeboundby

thetermsofthislicense.Forattribution,translations,adaptations,andpermissions,pleasereadtheprovisionsandtermsofuseat

/terms-use#openaccess

.

ThisCClicensedoesnotapplytonon-ADBcopyrightmaterialsinthispublication.Ifthematerialisattributedtoanothersource,pleasecontactthecopyrightownerorpublisherofthatsourceforpermissiontoreproduceit.

ADBcannotbeheldliableforanyclaimsthatariseasaresultofyouruseofthematerial.

Pleasecontactpubsmarketing@ifyouhavequestionsorcommentswithrespecttocontent,orifyouwishtoobtaincopyrightpermissionforyourintendedusethatdoesnotfallwithintheseterms,orforpermissiontousetheADBlogo.

CorrigendatoADBpublicationsmaybefoundat

/publications/corrigenda

.

Notes:

Inthispublication,“$”referstoUnitedStatesdollars.

CoverimagebyJamillahKnowles&Reset.TechAustralia/BetterImagesofAI/Socialmediacontent/CC-BY4.0.CoverdesignpreparedbyMikeCortes.

Contents

Forewordiv

Acknowledgmentsv

Abbreviationsvi

IntroductionfromtheInformationTechnologyDepartmentDirectorGeneralvii

Chapter1

ArtificialIntelligenceinDevelopmentOperations1

IntelligentConceptPaper—ProblemTreeGenerator2

DigitalAssistantforProjectAdministrationInstructions4

DigitalTwinsforSmartRoadConstruction6

ChatbotsasLifelinesforDomesticViolenceSurvivors8

DrivingSafeTransportationUsingDigitalSolutions11

RestoringPublicConfidenceonSafeTravels12

PredictiveModelforResettlementOutcomes14

Robo-Advisor16

LessonsLearnedinCognitiveSearch18

Chapter2

ArtificialIntelligenceinEconomics20

EconomicMonitoringUsingAlternativeDataSources21

UsingtheAutomaticIdentificationSystemtoGenerateInsightsonMaritimeActivities23

NowcastingtheEconomicImpactofCOVID-1925

EarthObservationData27

Chapter3

ArtificialIntelligenceinAdministration29

CodeofConductChatbot30

TechnicalAssistanceMessagingIntelligence31

MyADBRecruitmentIntelligence33

IntelligentIntegrityDueDiligencePlatform35

Chapter4

ADB’sArtificialIntelligenceBlogsandInsights37

AIandDevelopment—AgeofAbundanceorHugeDivide?

38

GenerativeArtificialIntelligenceandFinancialInclusion40

CouldChatGPTBeaGameChangerforInternationalDevelopment?40

SixWaystoPrepareTeachersinAsiaandthePacificfortheChatGPTEra41

FiveWaystoReaptheBenefitsoftheTechnologicalRevolution41

UsingMachineLearningonSatelliteImagestoMapPoverty42

ArtificialIntelligenceandMachineLearningintheTimeofCOVID-1942

DataandtheArtificialIntelligenceGoldRush:WhoWillWin?43

iii

Foreword

principlessuchasfairness,inclusiveness,transparency,

Strategy2030fromtheAsianDevelopmentBank

(ADB)explicitlystatesthatdigitaltransformation

isnecessarytomaketheorganizationstronger,

better,andfaster.Establishedin2018,theADB

fairuse,accuracy,andsecurity,ADBcanimprovetheeffectivenessandefficiencyofitsoperations,includingthesupportitprovidestodeveloping

membercountries(DMCs).

WiththerapidevolutionofAIandthetransition

fromtheDigitalInnovationSandboxtotheDigital

LearningLabs,ADBisprimedtodiscovernewand

innovativewaystousethisdigitaltechnology,notonlyforfuture-proofingbutalsotohelpDMCsadaptto

modernchallenges.Forexample,AI-poweredtoolsforclimatemodelingcanhelpgovernmentsidentifytheanticipatedimpactsofclimatechangeandtakeproactivemeasures.Similarly,AI-poweredtoolscanbeusedtooptimizefoodsupplychainsorpredict

diseaseoutbreaks.

Beforewelookintothefuture,wemustlearnfrom

ourpastexperiencestohelpusgrowandadapt.ThispublicationprovidesaglimpseofsomeoftheAI

workADBhasdoneinAsiaandthePacific.ItisworthnotingthatthebankdidnotjustexplorewaysinwhichAIcouldbeusedtoimproveinternalprocesses;rather,italsotestedhowthedigitaltechnologycouldbeusedtoaddressreal-worldchallengesintheareasofhealth,economy,transport,andsocialprotectionfacedby

DMCs.MostofthesewereimplementedthroughtheDigitalInnovationSandbox.

However,threetools—theCodeofConductChatbot,theChatbotsasLifelinesforDomesticViolence

DigitalInnovationSandboxisoneoftheprograms

thatemergedtoattainthisambitiousgoal.Itserves

asasafeandneutralfacilityforADBtotestand

reviewnewandemergingdigitaltechnologiesinfive

areas—artificialintelligence(AI),robotics,distributed

ledgertechnology,bigdata,andmixedreality—tosee

whetherthesehavethepotentialtomeetbusinessand

operationsneedsanddrivetheDigitalAgenda2030.

AIwasstillanemergingtechnologywhenthe

DigitalInnovationSandboxwasestablishedin2018.

Nonetheless,ADBsawitsvalueandpotentialin

helpingdelivertargets.Thus,theearliestinitiatives

thatweretestedthroughtheDigitalInnovation

SandboxwereAI-driven.Atthattime,thework

focusedonmachinelearning.Onesuchinitiativewas

Robo-Advisor,whichcouldanswerquestionsrelated

toADB’soperationsrelatedtogender,finance,and

climatechange,6yearsbeforethelaunchofChatGPT.

Theemergenceofdeeplearningopenedevenmore

opportunitiesforADBtotestAIinitswork,leading

toinitiativesthatusedvariousdatasourcestopredict

economictrends.Asofthistimeofwriting,ADB

isexploringgenerativeAI,whichisprojectedto

introduceanewdevelopmentera.

AIholdsalotofpotentialtochangetheworld,but

thesechangescangobothways.TheAIGovernance

Survivors,andtheAutomaticIdentificationSystemtoGenerateInsightsonMaritimeActivities—

werespearheadedbyotherADBdepartments,an

indicationofthegrowingroleofAIinthebank’swork.

Ihopethatthispublicationwillspurfurtherresearch

anddevelopmentonwaysinwhichAIandotherdigitaltechnologiescouldbeleveragedtofosterinclusiveandprosperousdevelopmentforthebenefitofall.

WorkingGroup,whichcomprisestheBudget,

Personnel,andManagementSystemsDepartment;

ClimateChangeandSustainableDevelopment

Department;DepartmentofCommunicationsand

KnowledgeManagement;InformationTechnology

Department;OfficeoftheAuditorGeneral;Officeof

theGeneralCounsel;OfficeofthePresident;Office

ofProfessionalEthicsandConduct;OfficeofRisk

Management;Procurement,Portfolio,andFinancial

ManagementDepartment;Strategy,Policy,and

PartnershipsDepartment;SectorsGroup;Southeast

AsiaDepartment;OfficeoftheVice-President

MuhammadEhsanKhan

DirectorGeneral,EastAsiaDepartmentand

IntelligenceGovernanceWorkingGroupChairAsianDevelopmentBank

(AdministrationandCorporateManagement);and

theoperationsdepartments,providesguidanceon

theethicalandresponsibleuseofAIacrossADB.

Byharnessingthetechnology,temperedbyguiding

iv

Acknowledgments

FinancialManagementDepartment;

?AlaysaTagumpayEscandor,publicmanagementofficer(Governance)atSectorsGroup(SG);

?ArunRamamurthy,principaltransportspecialistatSG;

?PawanKarki,principaltransportspecialistatSG;and

?JaeKyounKim,healthspecialistatSG.

SpecialthanksgotoITDdirectorgeneralStephanie

HungandEastAsiaDepartmentdirectorgeneralandArtificialIntelligence(AI)GovernanceWorkingGroupchairMuhammadEhsanKhanfortheirinvaluable

supportinthedevelopmentofthispublication.

Theteamthanksitspartners,theAsianInstituteofManagementandOracleLabs.Inaddition,

acknowledgmentisgiventotechnologyservice

providersthatworkedwithADBtodevelopthe

AI-poweredinitiativesfeaturedinthisknowledge

product:ORIS,NeuralMechanicsInc.,andYellow.ai.

CreditfortheimagesinthispublicationgoestoBetterImagesofAI,anonprofitcollaborationofvarious

individualsandnonprofitandacademicinstitutions

thatadvocatesforthecorrectrepresentationofAI,

consideringthatmanypubliclyavailableimagesoftenmisrepresentthetechnologyandreinforcenegativeorevenharmfulstereotypes.

ArtificialIntelligenceinAction:SelectedADBInitiatives

inAsiaandthePacificwasdevelopedbythe

InformationTechnologyDepartment(ITD)ofthe

AsianDevelopmentBank(ADB).

MarcLepage,principalinformationtechnology

specialistforTechnologyInnovationatITD,

conceptualizedandguidedthedevelopmentof

thisknowledgeproduct.OzzeirKhan,ITDdirector,

providedguidance.

MelanieSison,technicalwriter,preparedthecase

studieswithsupportfromthefollowingADB

colleagues,whoprovidedtechnicalinputand

feedbackonthedocumentationofinitiativesthat

involvedtheirrespectiveoffices:

?PierreDyens,principalhumanresourcespecialistattheBudget,Personnel,andManagementSystemsDepartment;

?VeronicaJoffre,seniorgenderandsocial

?developmentspecialistattheClimateChangeandSustainableDevelopmentDepartment;

?NellyDefo,seniorfinancialcontrol

?specialistattheController’sDepartment;

?PinkySerafica,seniorcommunicationsofficerattheDepartmentofCommunications

andKnowledgeManagement;

?AbdulAbiad,directoroftheEconomicResearchandDevelopmentImpactDepartment(ERDI);

?ElaineS.Tan,advisoratERDIandheadoftheStatisticsandDataInnovationUnit;

?HomerPagkalinawan,associateeconomicsofficeratERDI;

?MadhaviPundit,senioreconomistatERDI;

?MayaVijayaraghavan,principalevaluation

specialistattheImpactEvaluationDepartment;

?KadraSaeed,principalinformationtechnologyspecialist(DigitalProductsLead)atITD;

?ChristelAdamou,directoroftheOfficeofProfessionalEthicsandConduct;

?MariaLorenaCleto,socialdevelopmentspecialist(Resettlement)attheOfficeofSafeguards;

?FionaAlpe,principaltransactionsupportspecialistatthePrivateSectorOperationsDepartment;

?XiufengZhao,principalportfoliomanagementspecialistattheProcurement,Portfolio,and

v

Abbreviations

ADB

AI

AIS

AOI

COVID-19CTLA-TADBSCANDMC

AsianDevelopmentBank

artificialintelligence

AutomaticIdentificationSystem

areaofinterest

coronavirusdisease

Controller’sDepartmentLoanAdministrationDivision—TechnicalAssistanceSectionDensity-BasedSpatialClusteringofApplicationswithNoise

EIM

EOI

GDP

GPS

H3

ICP

IDD

IED

IT

ITD

km

LAR

MARIMVP

NCAVNLI

OPECPAI

PIA

POC

developingmembercountry

economicindicatormodel

eventofinterest

grossdomesticproduct

globalpositioningsystem

HexagonalHierarchicalGeospatialIndexingSystem

IntelligentConceptPaper

integrityduediligence

IndependentEvaluationDepartment

informationtechnology

InformationTechnologyDepartment

kilometer

landacquisitionandresettlement

MyADBRecruitmentIntelligence

minimumviableproduct

NationalCenterAgainstViolence(Mongolia)

NationalLegalInstitute(Mongolia)

OfficeofProfessionalEthicsandConduct

ProjectAdministrationInstructions

PPFDIntelligentAssistant

proofofconcept

PPFD

PSOD

RT-PCRSDG

TA

TAM

TAMI

TASU

Procurement,PortfolioandFinancialManagementDepartment

PrivateSectorOperationsDepartment

reversetranscriptionpolymerasechainreaction

SustainableDevelopmentGoal

technicalassistance

themeaugmentedmodel

TechnicalAssistanceMessagingIntelligence

TechnicalAssistanceSupervisingUnits

vi

Introductionfromthe

InformationTechnologyDepartmentDirectorGeneral

AttheInformationTechnologyDepartment,we

areexploringwaysinwhichAIcanbeleveragedtoredefinethewaywework,and,moreimportantly,craftsolutionsthatwouldhelpaddressthe

increasinglymultifacetedissuesfacedbyAsia

andthePacific,particularlythepoorestandmost

vulnerablecommunitiesintheregion.Wearenot

onlycollaboratingwithotherdepartmentsinADB

andwithstakeholdersinDMCs;wearealsopursuingandnurturingpartnershipswithindustryleadersandinnovativestartupstoachieveourgoalofresponsibleAIdevelopment,allowingusearlyaccesstotest

newtoolsandtechnologiestoseewhetherthey

arecompatiblewithADB’ssystemsandprocesses.

Collaborationalsooffersadditionalbenefitsinthe

contextofAIinnovations:byleveragingthecollectiveintelligenceandexperienceofexpertsfromvarious

fieldsandbackgrounds,itcanaccelerateprogress.ItcanalsohelpmitigatebiasesandethicalissuesinAIdevelopment.

TheworkwehavedoneinAIinthepastyearshasresultedintheaccumulationofknowledge,and

thispublicationprovidesuswithsnapshotsofsomeofwhatwehavelearnedinthelast6years.We

hopethatyoucanviewtheseusecasesthrougha

growth-mindsetlens.AIisstillinthenascentstagesofdevelopment,whichmakesthisthebesttimeforustotest,experimentwith,andlearndifferentwayswecanuseAIindifferentcontexts.Letuslearnandgrowtogether.

Thedevelopingmembercountries(DMCs)ofthe

AsianDevelopmentBank(ADB)faceincreasingly

challengingcircumstances.Evenastheypivottoward

recoveryafterbeingheavilyhitbythecoronavirus

disease(COVID-19),theyalsohavetocontendwith

climatechange,biodiversityloss,andotherissuesthat

threatennotonlytohinder,butreversethehard-

earneddevelopmentprogressthattheyhavestriven

forinthelastdecades.

Inthepastyears,ADB’srolehasevolvedfrombeing

afinancingpartnertoalsoservingasaknowledge

bankthatenablesthedevelopmentofDMCs.

Knowledgeisthegroundonwhichinnovationstands,

andinnovation,inturn,canaccelerateeconomic

growthandsocialprogress,especiallyinthecontext

ofknowledgeasapublicgood.Thisalsoimpliesthat

puttingminimalfocusonknowledgeandinnovations

canputnotjustthebank,butalsotheDMCsitserves,

atriskofbeingleftbehind.

Artificialintelligence(AI),particularlygenerativeAI,

whilenotasilverbullet,canbeapowerfultoolthat

wouldhelpputusontherightpathtoaddressthese

growingchallenges.Ithasthepotentialtoserveasa

keydriverofgrowthandinnovation.However,this

canonlybefullyrealizedifsufficientinvestmentsare

madenotonlytoeffectivelyharnessthetechnology

butalsotoempowerteamstobecomecompetent

inresponsiblyleveragingAItoaugmentdecisions

andenhanceproductivityandefficiency.Afterall,

ADB’sdigitaltransformationdoesnotjustrefertothe

hardwareandsoftware;thecapacityofitspersonnel

dictatesthespeedatwhichtechnologywillbe

adoptedbythebank.

StephanieKing-chungHung

ChiefInformationOfficerand

DirectorGeneral,InformationTechnologyDepartment

AsianDevelopmentBank

vii

ArtificialIntelligenceinDevelopmentOperations

CHAPTER1

ThisimagedepictsthecomplexityofAI:whatmaybeperceivedasasingleentityisactuallyseverallayersofalgorithmslinkedtogethertoformacohesivesystem.

ImagebyAntonGrabolle/BetterImagesofAI/AIArchitecture/CC-BY4.0

Asadevelopment,knowledge,andclimatebank,theAsianDevelopmentBank(ADB)strivesto

supportdevelopingmembercountries(DMCs)in

lessonsfromprojectevaluations.Italsofeaturesways

inwhichthetechnologywasusedintheareasofroad

design,health,transport,andsocialprotection.Inthe

future,newusecasesofAIwilllikelyemergeasthe

worlddiscoversdifferentwaysinwhichitcanbeused

toresolvecomplexissues.

AsiaandthePacificbyofferingknowledge,expertise,

andfinancing.Theregionfacesincreasinglycomplex

challengessuchasclimatechange,poverty,andfood

insecurity.Addressingtheseissuesrequiresforward

thinkingandnewsolutions.

ADBisexploringhowmoderntechnologiessuch

asartificialintelligence(AI)couldbeusedinits

operations.Thissectionshowsdifferentusecasesof

AIatvariousstagesoftheprojectcycle,fromthetime

theconceptpaperofaprojectiswrittentoextracting

WritingaconceptpaperisthefirststepADBtakes

tosupportDMCs.Todoso,writersneedtomanuallysiftthroughvariousinternaldocumentsandonline

materials;asaresult,asingleconceptpapercan

takebetweenseveralweeksandamonthtoprepare.Problemtrees,whichshowthecausesandeffectsofanissuethatADBseekstohelpresolve,canalone

take3to5daystocomplete.Eventhen,concept

papersareunlikelytobecomprehensivegiven

limitationsinfindingavailablereferences.Thus,ADBsoughttousetechnologytomakewritingconcept

papersfasterandmoreefficient.Thisinitiative

exploredusingAItoautomatethegenerationandpopulationofproblemtreesbasedonentered

searchqueries.

TheInformationTechnologyDepartment(ITD)engagedNeuralMechanicsInc.todevelopa

minimumviableproduct(MVP)ofthetoolthat

couldbeusedtoautomaticallygenerateproblem

trees.AIwasleveragedtoincreaseefficiencyand

reduceinconsistenciesinpreparingconceptpapers,particularlyproblemtrees,becauseitwasableto

processlargeamountsofdata.Thetool,whichwouldcometobeknownasIntelligentConceptPaper

(ICP),wastrainedonsimilaritymatching,enabling

ittofiltersentencesthataresemanticallysimilarto

textsfrompreviousproblemtrees.Aneuralnetworkmodel,whichisasubsetofmachinelearningthatcanidentifyrelationshipsbasedonattributedweights,

wasthenusedtoidentifycausaleventpairsfromthefilteredsentences.

2

IntelligentConceptPaper—ProblemTreeGenerator

SUMMARY

Whatitis:

TheIntelligentConceptPaperwasanartificialintelligence-powered

systemthatautomaticallygeneratedproblemtrees,shorteningthetimeneededtoprepareconceptpapers.

Startofimplementation:

2019

Implementationpartner:

EastAsiaDepartment

Technologyserviceprovider:

NeuralMechanicsInc.

InlinewithADB’soperationalapproaches:

?Promotingdigitaldevelopmentandinnovativetechnologies

?Applyingdifferentiatedapproaches

?Deliveringintegratedsolutions

Usersenteredtheirownproblemstatementsor

used/editedthesuggestionsthatwereauto-generatedbasedontheirenteredkeywords.Fromthere,theyhad

theoptiontousethesuggestedcausesandeffects

ThesourcesofinformationweredocumentsfromADB’sEastAsiaDepartmentandtheWorldBank,aswellasXinhuanewsarticles.Thesewerefirst

convertedtoareadableformattoenablethedesignedwebscrapertocrawlthroughthesedocuments.

Thesewereannotatedtomakesearchingeasier

fortheengine.Theannotatedmaterialswerethen

storedinthecloud.Anappwasdevelopedtorecoverrelevantinformationbasedonthekeywordsenteredbyusers.

orwritetheirown.Itwaspossibletoaddmultipleresponsesforbothcausesandeffects,aswellassub-causesandsub-effects,ifuserswantedto

providemorespecificinformation.

Aclick-and-dropfeaturewasalsoincorporatedintoICP’sdesigntobuildtheproblemtreerightonthe

platform.Userswereabletosharethiswithothersfortheirreviewandinput.

ArtificialIntelligenceinAction

ThesefeaturesallowedICPtogenerateresultswithin

seconds,significantlycuttingdownthetimeneeded

toresearchinformationandproducetheproblem

treediagram.Thetoolalsoallowedvarioususersto

collaborateonthesameproblemtreeanddocument,

makingworkmoreefficient.

WorkonICPbeganin2019andwascompletedin

April2020.Whilethistoolisnolongerinuse,the

initiativeshowedthatitwaspossibletodevelopan

AI-drivenplatformthatbuiltproblemtrees.Thetool

markedvariousfirstsforADB:thefirsthomegrown

AIforADBoperations,thefirstonedevelopedthat

leveragedthewealthofknowledgefromprevious

worksofADBwithinthedevelopmentlandscape,and

thefirsttoharnessbothtechnologyandhumaninputs

toscanthroughkeymaterialstoidentifyproblemsin

ADB’sDMCs.

ADB-wideconsultationsand/orfocusgroup

discussionsmaybeconductedtoexplorehowlessons

fromthetooldevelopmentcanbeusedinthefuture,

includinginwritingothersectionsofaconceptpaper

suchastheDesignandMonitoringFrameworks.

ArtificialIntelligenceinDevelopmentOperations3

ProjectAdministrationInstructions(PAIs)areamongthemostaccessedmaterialsontheADBwebsite,withover20,000viewsanddownloadsyearly.TheseveralseriesofPAIsprovideinformationonthepolicies

andproceduresthatshouldbefollowedthroughouttheentireprojectcycleofADB-financedloanandtechnicalassistance(TA)projects.1

Usersspentfromafewminutesuptoseveralhoursfindingtheinformationtheyneeded.Tomakeit

easiertofindkeyinstructionsandguidelinesinPAIs,Yellow.ai,thencalledYellowMessenger,wasengagedtodevelopadigitalassistant.Theintentwasto

haveatoolthatcouldquicklyprovideuserswiththeinformationtheyneedwithouthavingtomanuallysiftthroughvariousdocumentsontheADBwebsite.

WorkontheMVPtookplacebetweenJune

andSeptember2020.Keypersonnelfromthe

Procurement,PortfolioandFinancialManagement

Department(PPFD)testeditbetweenSeptember

andOctober2020,andcommentsandfeedbackwereconsideredinthedevelopmentofthedigitalassistant,whichwasofficiallylaunchedinOctober2020.The

digitalassistant,whichcametobeknownasPPFDIntelligentAssistant(PIA),continuestobeused

in2024.

ThevirtualassistantispoweredbyAI,enablingit

tointeractwithusersinaconversationalmanner

andgenerateresultsrelevanttoqueries.Traditionalkeywordsearchesreliedonusers’familiarity

DigitalAssistantforProjectAdministrationInstructions

SUMMARY

Whatitis:

TheDigitalAssistantforProject

AdministrationInstructionsmakesfindinginformationonprojectimplementation

easierforADBclientsandstaff.

Startofimplementation:

2020

Implementationpartner:

Procurement,PortfolioandFinancialManagementDepartment

Technologyserviceprovider:

Yellow.ai

InlinewithADB’soperationalapproach:

?Promotingdigitaldevelopmentandinnovativetechnologies

informationtousersbecausePAIdocumentsandtechnicalguidancenotesareupdatedonanadhocbasis.OfflinesupportfromPPFDpersonnelis

alsoavailableincasePIAisunabletoprovidetherequiredinformation.

withspecificterms.Incontrast,AIwithnatural

languageprocessingcanbetrainedtogenerate

resultsaccordingtosearch