DUKLEYAI LIVEInvestor view ↗︎

THE INTELLIGENCE LAYER OF DUKLEY

The company that
can see, control
and analyze itself.

AI native.

Connecting company data, operational signals and human decisions into one governed intelligence system.

12 DEPARTMENTS · DEMO50+ OPERATIONAL AREAS2 PRIORITY PILOTS · DEMO
CHECKING VERIFIED DATA

THE IDEA

FROM DATA TO DECISION.
FROM DECISION TO EXPERIENCE.

Choose a step to see how the system behaves in practice.

HOW IT WORKS

Connect

Dukley AI begins by connecting authorised corporate sources without replacing the systems teams already use.

DUKLEY EXAMPLEA new technical request, an updated Drive file and a CRM change become visible in the same controlled layer.

WHAT IS LIVE NOW

VERIFIED, NOT PROJECTED.

Current programme state. No potential result is presented as achieved.

PROGRAMME DATA FEEDCHECKINGNo verified snapshot has been received yet. Demo values are clearly labelled.

Data foundation

CheckingConnecting to the verified programme feed

Collection layer

CheckingConnecting to the verified programme feed

Control Center

CheckingConnecting to the verified programme feed

Priority pilots

CheckingConnecting to the verified programme feed
SYSTEM READINESS5 TRACKED LAYERS
2Active

Data foundation and Control Center

LIVE SYSTEM MAP

ONE SYSTEM.
FIVE CONNECTED LAYERS.

Select a layer to see its role and verified state.

LAYER 01ACTIVE

One governed data foundation

Google Workspace is the corporate core. Root workflow folders are distributed and document migration is in progress.

Google WorkspaceRead-only collectionMetadata index

ENTERPRISE ARCHITECTURE MIND MAP

ONE OPERATING LAYER.
EVERY CONNECTION VISIBLE.

Select a branch to explore how data becomes governed intelligence, department action and a visible Dukley experience.

COMPANY CONTROL PLANEDUKLEY
AI OS
Control API · Registry · State · Audit
LAYER 01

Data & Knowledge

Google Workspace, operating systems and structured registries provide one authorised foundation. Information remains in its source systems while Dukley AI builds controlled context around it.

01Google Workspace02Operating systems03PostgreSQL registry04Company Memory
OPERATING FLOW
01Observe
02Understand
03Recommend
04Human decides
05Act
06Check result

AI SOLUTIONS PORTFOLIO

REAL SOLUTIONS.
CLEAR OWNERSHIP.

A solution is tracked as a business capability, not just a workflow.

DATA / 01IN BUILD

Corporate Data Pulse

OWNERDukley AIENVIRONMENTPilot↗︎
OPS / 02PILOT

Technical Service Assistant

OWNERTechnical ServiceENVIRONMENTPilot↗︎
MKT / 03PREPARED

Marketing Intelligence

OWNERMarketingENVIRONMENTPrepared↗︎
OPS / 04LIVE TEST

Kolašin Service Bot

OWNERKolašinENVIRONMENTTest↗︎

FUTURE IDEAS LAB

THE VISIBLE FUTURE
OF DUKLEY AI.

Hover or tap an idea to see the explanation and a concrete Dukley example.

PRIORITY CONCEPT01 / 11
DUKLEY AI
•••
GOOD EVENING, ALEX

Your stay, quietly coordinated.

RESIDENCE READY
TRANSFER18:40 · Confirmed
DINNER20:30 · Seafront
HOMEAISERVICE

AI Concierge

THE IDEA

A multilingual concierge combines property knowledge, guest context and approved service actions in one elegant conversation.

EXAMPLE AT DUKLEY

Before arrival, it proposes a transfer and dinner based on timing and preferences. During the stay, it coordinates requests without making the guest repeat details.

Strategic fitInvestor visibilityData readiness

BUSINESS IMPACT

MEASURE FIRST.
SCALE SECOND.

The pilot KPIs are defined. Baselines and results will appear only when collected.

IMPACT MEASUREMENT READINESS
DefinedBaselineMeasured
WHAT WE WILL MEASURE

Classification accuracy

Share of requests correctly classified by the pilot compared with human review.

WHY IT MATTERSShows whether routing can become reliably faster without losing control.
PROGRAM MOMENTUMCHRONOLOGICAL VIEW · AUGUST 2026

10 AUG

Workflow foundation distributedCorporate root folders shared with managers.DONE

11 AUG

MVP Control Center activatedSolution ownership, status and criticality brought into one view.DONE

20 AUG

Single-system scope confirmedData and AI solution visibility aligned inside one architecture.CURRENT

26 AUG

Pilot validation gateRepeat sync and action-brief readiness reviewed before scale.NEXT

ECONOMICS & ROI OF TRANSFORMATION

A BUSINESS CASE
FOR EVERY AI DECISION.

Dukley AI OS is deployed only when the financial logic is explicit, measurable and ready to be tested against pilot evidence.

DECISION LOGICAS-IS COSTVSAI TCO+ TRANSITION & CONTROL
BASELINE01 / 02

As-Is process cost

We establish the full cost of the process before automation so that value is measured against a real operating baseline.

01Manual labour02Cost of errors03Time loss
DUKLEY EXAMPLEFor a technical request: staff time spent reading, clarifying and routing the request, plus delays, repeat visits and correction work.
OPEX OPTIMISATION TARGET20TO30%

Ambition for qualifying pilot processes, subject to baseline and validation.

FORECAST PAYBACK WINDOW6TO12MONTHS

Expected payback period for approved solutions, validated case by case before scale.

PHASED LAUNCH & ROADMAP

VALUE FIRST.
COMPLEXITY IN CONTROL.

We do not launch every solution at once. The AI Opportunity Matrix prioritises each case by business value and technical complexity.

AI OPPORTUNITY MATRIXSELECT A PHASE
BUSINESS VALUE ↑TECHNICAL COMPLEXITY →
PRIORITISE
STRATEGIC BETS
DEFER
ENABLE
PHASE 01High business value · lower complexity

Foundation that proves value

Start with controlled use cases that establish shared memory and create measurable operational evidence quickly.

DELIVERABLESDukley Company MemoryTechnical Service request-processing pilot

SECURITY & CONTROL · AI GOVERNANCE

TRUST IS DESIGNED
INTO THE SYSTEM.

Investor and guest data security is our priority. A strict corporate AI Policy and Human-AI operating model govern every solution.

01HUMAN APPROVAL

Human-in-the-loop

AI proposes decisions and automates routine work, while critical actions—including budget approvals and final reports—always require human approval.

02PROTECTED ENVIRONMENT

Confidentiality

Commercial information and guests’ personal data are processed inside a closed, protected environment with role-based access.

03CONTINUOUS AUDIT

Monitoring

AI outputs are continuously audited to detect hallucinations, errors and quality drift before they affect a decision or experience.

RISK CONTROL MODEL

Select a risk level to see its control rule and a practical example.

RISK LEVEL L1

Informational

AI reads approved information and produces low-risk summaries.

DUKLEY EXAMPLESummarising a public property description for an internal brief.
SSO / MFAROLE ACCESSAUDITHUMAN APPROVAL

LIVE INVESTOR CONTROL ROOM

FROM VISION
TO INVESTMENT EVIDENCE.

● VERIFIED & TARGETS SEPARATED
PROPOSED MOBILE INVESTOR EXPERIENCE

The whole transformation, in one clear view.

A private investor interface can combine verified delivery status, ROI targets, governance gates and the next management decision—without exposing operational or personal data.

01Live programme pulse02ROI and evidence03Decision-ready briefs
INTERFACE CONCEPT · READY FOR LIVE DATA CONNECTION
DUKLEY AI
WEEK 34
INVESTOR OVERVIEW

Good morning.
Here is what moved.

2/5LAYERS
ACTIVE
SYSTEM READINESSFoundation
is taking shape.
Verified · 20 Aug 2026
OPEX TARGET20–30%pilot processes
PAYBACK6–12months forecast
LIVE SIGNALn8n collection pilot is operating.Next gate: data baseline validation
PULSEROIROADMAPCONTROL
LIVE COLLECTION1

n8n pilot operating

PRIORITY PILOTS2

Marketing · Technical Service

DEFINED PILOT KPIs5

Baselines next

CHOOSE AN INVESTMENT LENSWHAT SHOULD DUKLEY AI PROVE?
INVESTOR QUESTION

Will guests feel a meaningful difference?

Recurring friction, personalisation opportunities and service actions that can be verified—not just described.

WEEKLY INTELLIGENCE BRIEF14–20 AUG 2026

The operating foundation is becoming visible.

01CompletedSingle Control Center scope and investor experience aligned.
02In motionWorkspace migration, repeat collection and observation layer.
03Next gateValidate the data foundation before scaling agents.
90-DAY INVESTMENT PATHEVERY PHASE ENDS WITH A DECISION
01ENTRY

Readiness check

Value, data access, technical effort, risk and ownership are scored before build.

02PROOF

Controlled pilot

A narrow workflow runs with KPI, access, logs and human approval from day one.

03EVIDENCE

ROI evidence

As-Is cost, AI TCO and measured operational change create the business case.

04DECISION

Scale / No-scale

Only useful, safe and economically justified solutions enter the wider portfolio.