The AI Lab builds tools, programmes, and infrastructure that help every team across the
Group work smarter. Over 1,600 employees already use our AI platform,
ALIA, which routes tasks across 16 AI models to match each person with
the right tool for the job. We are not replacing people; we are giving them superpowers.
What We Do
01
Build AI Tools
We design and deploy AI-powered solutions that solve real operational problems, from customer care automation to warehouse planning.
02
Upskill Our People
Through ALIA, scholarships, and the AI Sandbox, we help every employee grow their AI capabilities regardless of technical background.
03
Measure Impact
We track how AI changes productivity and cost structures, making sure investments translate into real business value, not just adoption numbers.
04
Drive Culture
AI only works when people trust and use it. We run hackathons, communities of practice, and champions programmes to build that culture from the ground up.
ALIA: The Group's AI Platform
Your AI assistant, built for Chalhoub
ALIA is the Group's generative AI platform. Available through web and Microsoft Teams,
it gives every back-office employee access to a suite of AI models through a single,
secure interface.
Unlike generic tools, ALIA is built around how our people actually work, with a model router
that selects the right AI for each task: drafting content, analysing data,
generating images, or reasoning through complex problems.
Intelligent model routing across 16 AI models, selecting the right model for each task automatically
Accessible via web browser and Microsoft Teams, no new apps to install
Secure and private: conversations are not used to train external models
Tiered access that grows with your skills, from basic chat to advanced agent workflows
Built-in usage analytics that track your AI maturity journey
Your AI Journey
Everyone starts somewhere. The AI maturity ladder is a path, not a test. Each level unlocks
new capabilities and tools. Move at your own pace.
1
Aware
Getting Started
You know AI exists and have heard of ALIA. You may have tried it once or twice. This is where most people begin.
Unlocks: ALIA basic access
2
Basic
Building Habits
You use ALIA regularly for straightforward tasks: drafting emails, summarising documents, brainstorming ideas. AI is becoming part of your routine.
Unlocks: All standard ALIA features
3
Regular
Working Smarter
You choose the right AI approach for different tasks. You write effective prompts, use templates, and share what works with your team.
Unlocks: Advanced models + custom templates
4
Advanced
Leading Change
You identify new AI use cases for your team, build workflows, and mentor others. You are a go-to person for AI questions in your department.
Unlocks: GenAI Scholarship Programme
5
Power
Building the Future
You prototype AI solutions, work with APIs, and collaborate directly with the AI Lab to bring new tools to life. You are shaping how the Group uses AI.
Unlocks: AI Sandbox, a hands-on building environment
Our Initiatives
KAIA handles routine customer enquiries across channels, resolving common questions instantly and routing complex cases to the right human agent with full context attached.
The system learns from resolution patterns to improve over time, reducing repeat contacts and ensuring consistent quality regardless of channel or time of day.
Faster response times, consistent quality across channels
How it works
Customer sends a messageEmail, chat, or social media. Any channel.
KAIA classifies the intentAI determines what the customer needs and whether it can be resolved automatically.
Routine queries: auto-resolvedOrder status, return policy, store hours. Answered instantly from verified sources.
Complex cases: routed with contextThe agent receives the full conversation, customer history, and a suggested resolution.
WATSON analyses customer behaviour patterns to generate campaign strategies, segment audiences, and recommend the right message at the right time.
Instead of manual audience building and guesswork, marketing teams get data-driven recommendations that compress campaign setup from days to hours.
Smarter campaigns, set up in hours instead of days
How it works
Ingest customer dataPurchase history, browsing behaviour, engagement signals across channels.
Identify behaviour patternsML models cluster customers by lifecycle stage, preferences, and purchase likelihood.
Generate campaign recommendationsAudience segments, messaging angles, timing, and channel mix, tailored to each cohort.
Marketing team reviews and launchesHuman approval before anything goes live. The team stays in control.
OPTIVAULT forecasts demand patterns across distribution centres and generates shift allocations that match staffing levels to expected workload.
The model accounts for seasonality, promotional calendars, and historical throughput to prevent both overstaffing and bottlenecks.
Model staffing requirementsML translates volume forecasts into required headcount by role and skill level.
Generate shift allocationsOptimised schedules that balance efficiency, compliance, and employee preferences.
Continuous recalibrationActuals feed back into the model. Accuracy improves with every cycle.
Product data arrives from hundreds of brands in different formats, with inconsistent attributes, missing fields, and varying quality. PGR uses AI to enrich, standardise, and validate this data into a single golden record per product.
The result is cleaner catalogues, better search and filtering on-site, and less manual cleanup for merchandising teams.
One clean source of truth for product data across every brand
How it works
Raw product data ingestedBrand feeds, linesheets, and PIM exports in varying formats and completeness.
AI extracts and enriches attributesVision models read product images. Language models parse descriptions. Missing fields are inferred and validated.
Standardised to group taxonomyColour, material, occasion, fit, and category mapped to a consistent structure across brands.
Golden record publishedClean, enriched product data flows into PIM, search, and merchandising systems.
As consumers increasingly use AI-powered search (ChatGPT, Perplexity, Google AI Overviews), brands need to surface accurately in generated answers, not just traditional search results.
GEO ensures our brands are correctly represented when AI engines answer questions about luxury products, stores, and services in the region.
Our brands surface accurately in AI-powered search and answer engines
How it works
Audit current AI visibilityHow do AI engines currently describe our brands? Where are they wrong, incomplete, or missing?
Optimise structured contentSchema markup, authoritative content, and data feeds tuned for AI comprehension.
Monitor answer qualityAutomated checks across major AI platforms to catch inaccuracies and track improvements.
Iterate and expandNew brand launches, seasonal campaigns, and market changes trigger content refreshes.
The AI Sandbox is the top tier of the maturity journey. Power users get access to APIs, development tools, and direct mentorship from the AI Lab to prototype and test AI solutions with real data.
The goal is to remove the bottleneck between having an idea and testing whether it works, without waiting for an IT project to be scoped and funded.
From idea to working prototype without waiting for IT
How it works
Submit a use caseDescribe the problem you want to solve. The AI Lab reviews feasibility and data requirements.
Get sandbox accessAPI keys, development environment, sample datasets, and a mentor from the AI Lab.
Build and test your prototypeIterate quickly with real tools. No procurement, no infrastructure requests.
Evaluate and scalePromising prototypes get reviewed for production. The best ideas become full initiatives.
How We Measure Impact
Adoption is not impact. We track AI value through a four-layer framework that connects platform
usage all the way to business outcomes.
1
Consumption
Are people using the AI tools?
Adoption metrics: logins, sessions, onboarding rates. Necessary but not sufficient. Most organisations stop here.
2
Work
What work is AI actually doing?
Task-level measurement. What tasks were completed with AI assistance, how quickly, and at what quality. This is where productivity gains live.
3
Outcomes
Are business results improving?
Operational metrics: cost per unit of output, throughput, error rates. The translation from "we used AI" to "it made a measurable difference."
4
Business Impact
Does it show up in the P&L?
The ultimate test. Revenue growth, cost reduction, margin improvement. Where AI investment becomes business value that leadership can see and act on.
Get Involved
AI works better when everyone is part of it
Whether you have an idea for a new use case, want to join the AI community,
or just want to learn more about what ALIA can do for your team, we would love to hear from you.