Transforming Thomas Cook’s static tour packages into an AI-powered, conversational travel planning experience.
Tacy AI Ecosystem
Role
Lead UX/UI Designer & Product Strategist
Company
Atirath Technologies x Thomas Cook India
Tools
Figma
FigJam
REACT
Notion
Miro
Launched: January 2025
Atirath Website
Starting Point: Legacy Experience
Thomas Cook’s digital presence was built around static, pre-defined packages and traditional booking flows. Users had to jump between:
01
Marketing pages for destination research
02
PDF Itineraries for package details
03
Separate forms or call centers for modifications and bookings
There was no single, continuous experience where a traveler could explore, customize, and book a trip without starting over each time they changed their mind.
Project Overview
TravBridge (internally, part of the Tacy AI Ecosystem) is a Generative AI–powered travel assistant that helps Thomas Cook customers plan vacations from scratch or using curated packages, all inside a conversational interface.
The system contains
An AI chatbot for destination discovery and trip planning
A dynamic package builder that updates itineraries in real time
Fixed Thomas Cook–inspired packages for fast, low-effort bookings
Support for revisiting old chats, saved packages, and sharing itineraries with co-travellers
Outcomes at a glance
Designed an end-to-end chat-first UX for travel research, package creation, and booking.
Defined 5+ core workflows (explore destinations, dynamic packages, fixed packages, My Chats, Packages, sharing).
Created a visual design system aligned with Thomas Cook (blue / yellow / white) and a desktop-first product UI ready for engineering handoff.
Established itinerary rules & constraints so AI-generated packages remain realistic and bookable.
The Challenge
How do you let users:
Talk to a chatbot like a travel agent
Customize trips endlessly
Still keep the structure, reliability, and compliance of Thomas Cook packages tours
Key Challenge 1
Designing a conversation model that feels natural but still collects all required trip data (destinations, dates, duration, travellers, budget, preferences).
Key Challenge 2
Balancing full customization (dynamic packages) with zero-effort options (Thomas Cook–inspired fixed packages).
Key Challenge 3
Ensuring continuity so users can pause and come back later without losing any planning work (My Chats, Packages tab).
Key Challenge 4
Encoding itinerary logic (arrival, departure, hotels, activities, transport) so the AI never creates impossible schedules.
Design Goals
01
Gen Z-friendly, AI-native UX
Make the chatbot the primary interface, not an add-on.
02
Single Continuos Journey
From Vacation inspiration to travel package customization to booking the travel package without context loss.
03
Two planning modes
Effortless curated packages vs. deep customization, both within the same UI.
04
Safe, realistic itineraries
Encoded rules so the AI respects travel constraints and package standards.
05
Scalable Design system
Flows and UI components that Thomas Cook can reuse for future AI features (visa rules, forex, etc.).
Ecosystem Map: Core Functionalities
From the design brief, I grouped TravBridge into five core pillars:
AI -Powered Destination Exploration
Dynamic Package Creation (Fully Customizable)
Thomas-Cook inspired Fixed Packages
My Chats – Revisit & Resume Conversations
Packages Hub – Saved & Booked Packages + Sharing
Each pillar became a dedicated workflow with its own triggers, UI states, and chatbot tone.
Key Workflows
01
Explore Destinations via Chat
Users start a new conversation from the AI Chat Helper / “New Chat” entry points.
The chatbot asks about budget, travel style, and interests (adventure, nightlife, off-beat, nature).
It returns short, scannable cards for each destination: best season, key attractions, visa conditions, and typical budgets.
Follow-up questions refine recommendations instead of sending users to new pages.
Goal: Replace scattered research with one focused, conversational flow.
02
Create Dynamic & Personalized Packages
This is the heart of the Tacy / TravBridge experience.
The chatbot offers the customer whether they want to create a fully personalized package or explore Thomas Cook packages first.
For a dynamic package, A persistent right-hand pane opens with an editable package shell where the customers can edit key properties of the package and personalize it according to their liking.
The itinerary generated by AI is bound to rules: Arrival and Departure blocks, Accommodation per night, Activities, Food, and Transport types.
Goal: Give travelers full control to shape their itinerary while ensuring the AI maintains structure, accuracy, and bookability.
03
Thomas-Cook Inspired Fixed Packages
The chatbot collects various key preferences from the customer and TravBridge surfaces pre-designed Thomas Cook packages with fixed itineraries.
Each packages shows various details to the customer like Duration, key cities, trip theme, inclusions, exclusions, visa conditions, and terms & conditions.
Customers cannot modify these itineraries but they can simply book the package with in a few simple clicks in the AI ecosystem’s interface.
Goal: Offer effortless, ready-to-book packages for users who prefer fast decisions over detailed customization.
04
My Chats: Continuity in planning
Since Trip planning happens over days or weeks, Every conversation is stored in a My Chats tab with filters inspired by Airbnb’s filter patterns (destination, date, status).
Reopening a chat restores the full message history and all the packages that have been discussed in the thread.
The chatbot can resurface older packages or suggest new package options if preferences have changed since that chat.
Goal: Let users return to any trip plan at any time without losing context, progress, or previously generated packages.
05
Packages Hub: Saved, Booked, and Shared
A dedicated Packages tab becomes the memory of the ecosystem.
The Packages hub shows the saved/booked packages whether it may be Dynamic or Thomas-Cook inspired packages to the customer at point of them they require it.
Customers can re-open a dynamic or fixed package and continue editing or viewing it before booking the package.
Customers can re-open a dynamic package and continue editing it before booking the package.
Goal: Create a central memory system where users can manage, re-open, and share all their packages effortlessly.
Itinerary Model & Rules
To prevent the AI from generating impossible trips, I defined a block-based itinerary system. Each day is made of blocks tagged as:
Image 1
Dynamic Package Itinerary Example
Arrival
First block of Day 1
Departure
Last block on last day
Accomodation
Check in/Check out into hotels and Overnight Stays
Activity
Attractions, Tours, and Experiences
Food
Restaurants and Meal Stops
Transport
Inter-city or Inter-country travel
This model keeps the UX clean while giving engineering a clear schema for AI-generated itineraries.
Visual Design System
Following Thomas Cook’s brand, I created a modern, AI-native UI layer:
Color Rationale
Voyage Blue (#0056B3)
Symbolizes trust, reliability, and forward momentum. As the primary brand color, it anchors the interface with a confident, travel-ready tone and reinforces Thomas Cook’s legacy of dependable service.
Sunshine Gold (#FFC107)
Represents energy, warmth, and moments of delight. Used as the accent color, it highlights key actions, important states, and interactive elements—guiding users naturally without overwhelming the interface.
Cloud Mist (#D9E6F2)
A soft, airy neutral that provides clarity and contrast throughout the UI. Ideal for panels, backgrounds, and layered surfaces, helping the conversational and itinerary elements feel clean, modern, and easy to scan.
Type Rationale
Aa
Bold Weight for Section Titles
Poppins
Regular Weight for Body and Chat text to keep conversations readable
Layout - Desktop first Responsive Grid
Left: Navigation & quick access to My Chats / Packages
Center: Chat conversation as the primary canvas
Right: Contextual pane for packages, filters, and itinerary editing
From Flows to UI (Process)
1
Identity & Design Brief - I formalized the problem space, goals, and competitive landscape (Ask Layla, PickYourTrail, Airbnb Experiences).
2
Workflow Mapping - Defined the six key workflows and their triggers, states, and success criteria (Destination exploration, Dynamic packages, Fixed packages, My Chats, Packages tab, Sharing).
3
Itinerary Rules & Edge Cases - Codified what the AI can and cannot change inside an itinerary.
4
Wireframes & Conversation Flows - Designed message sequences for how the chatbot collects preferences, announces when it has “all key details”, and then directly shows relevant packages instead of over-explaining.
5
High-Fidelity UI & Handoff - Built componentized Figma layouts for chat, panes, filters, and package cards, aligning with Atirath’s engineering stack (React).
Learnings & Impact
01
Designing for Conversations, Not Screens
Shifting from page-based navigation to dialogue-driven flows meant designing every question, response, and confirmation as a UX artifact, not just the UI container around it.
02
Guardrails Make AI Useful
The itinerary rules and non-deletable structural blocks turned “AI magic” into reliable, bookable trips instead of random suggestions.
03
Continuity = Trust
“My Chats” and “Packages” turned AI from a one-off gimmick into a long-term planning companion. Users always feel like the system “remembers” them.
04
Bridging UX & Front-End
Because the flows and visual system were designed with React implementation in mind, the project now serves as a template for future AI products within Thomas Cook and Atirath’s broader Tacy ecosystem.
Transforming Thomas Cook’s static tour packages into an AI-powered, conversational travel planning experience.
Tacy AI Ecosystem
Role
Lead UX/UI Designer & Product Strategist
Company
Atirath Technologies x Thomas Cook India
Tools
Figma
FigJam
REACT
Notion
Miro
Launched: January 2025
Atirath Website
Starting Point: Legacy Experience
Thomas Cook’s digital presence was built around static, pre-defined packages and traditional booking flows. Users had to jump between:
01
Marketing pages for destination research
02
PDF Itineraries for package details
03
Separate forms or call centers for modifications and bookings
There was no single, continuous experience where a traveler could explore, customize, and book a trip without starting over each time they changed their mind.
Project Overview
TravBridge (internally, part of the Tacy AI Ecosystem) is a Generative AI–powered travel assistant that helps Thomas Cook customers plan vacations from scratch or using curated packages, all inside a conversational interface.
The system contains
An AI chatbot for destination discovery and trip planning
A dynamic package builder that updates itineraries in real time
Fixed Thomas Cook–inspired packages for fast, low-effort bookings
Support for revisiting old chats, saved packages, and sharing itineraries with co-travellers
Outcomes at a glance
Designed an end-to-end chat-first UX for travel research, package creation, and booking.
Defined 5+ core workflows (explore destinations, dynamic packages, fixed packages, My Chats, Packages, sharing).
Created a visual design system aligned with Thomas Cook (blue / yellow / white) and a desktop-first product UI ready for engineering handoff.
Established itinerary rules & constraints so AI-generated packages remain realistic and bookable.
The Challenge
How do you let users:
Talk to a chatbot like a travel agent
Customize trips endlessly
Still keep the structure, reliability, and compliance of Thomas Cook packages tours
Key Challenge 1
Designing a conversation model that feels natural but still collects all required trip data (destinations, dates, duration, travellers, budget, preferences).
Key Challenge 2
Balancing full customization (dynamic packages) with zero-effort options (Thomas Cook–inspired fixed packages).
Key Challenge 3
Ensuring continuity so users can pause and come back later without losing any planning work (My Chats, Packages tab).
Key Challenge 4
Encoding itinerary logic (arrival, departure, hotels, activities, transport) so the AI never creates impossible schedules.
Design Goals
01
Gen Z-friendly, AI-native UX
Make the chatbot the primary interface, not an add-on.
02
Single Continuos Journey
From Vacation inspiration to travel package customization to booking the travel package without context loss.
03
Two planning modes
Effortless curated packages vs. deep customization, both within the same UI.
04
Safe, realistic itineraries
Encoded rules so the AI respects travel constraints and package standards.
05
Scalable Design system
Flows and UI components that Thomas Cook can reuse for future AI features (visa rules, forex, etc.).
Ecosystem Map: Core Functionalities
From the design brief, I grouped TravBridge into five core pillars:
AI -Powered Destination Exploration
Dynamic Package Creation (Fully Customizable)
Thomas-Cook inspired Fixed Packages
My Chats – Revisit & Resume Conversations
Packages Hub – Saved & Booked Packages + Sharing
Each pillar became a dedicated workflow with its own triggers, UI states, and chatbot tone.
Key Workflows
01
Explore Destinations via Chat
Users start a new conversation from the AI Chat Helper / “New Chat” entry points.
The chatbot asks about budget, travel style, and interests (adventure, nightlife, off-beat, nature).
It returns short, scannable cards for each destination: best season, key attractions, visa conditions, and typical budgets.
Follow-up questions refine recommendations instead of sending users to new pages.
Goal: Replace scattered research with one focused, conversational flow.
02
Create Dynamic & Personalized Packages
This is the heart of the Tacy / TravBridge experience.
The chatbot offers the customer whether they want to create a fully personalized package or explore Thomas Cook packages first.
For a dynamic package, A persistent right-hand pane opens with an editable package shell where the customers can edit key properties of the package and personalize it according to their liking.
The itinerary generated by AI is bound to rules: Arrival and Departure blocks, Accommodation per night, Activities, Food, and Transport types.
Goal: Give travelers full control to shape their itinerary while ensuring the AI maintains structure, accuracy, and bookability.
03
Thomas-Cook Inspired Fixed Packages
The chatbot collects various key preferences from the customer and TravBridge surfaces pre-designed Thomas Cook packages with fixed itineraries.
Each packages shows various details to the customer like Duration, key cities, trip theme, inclusions, exclusions, visa conditions, and terms & conditions.
Customers cannot modify these itineraries but they can simply book the package with in a few simple clicks in the AI ecosystem’s interface.
Goal: Offer effortless, ready-to-book packages for users who prefer fast decisions over detailed customization.
04
My Chats: Continuity in planning
Since Trip planning happens over days or weeks, Every conversation is stored in a My Chats tab with filters inspired by Airbnb’s filter patterns (destination, date, status).
Reopening a chat restores the full message history and all the packages that have been discussed in the thread.
The chatbot can resurface older packages or suggest new package options if preferences have changed since that chat.
Goal: Let users return to any trip plan at any time without losing context, progress, or previously generated packages.
05
Packages Hub: Saved, Booked, and Shared
A dedicated Packages tab becomes the memory of the ecosystem.
The Packages hub shows the saved/booked packages whether it may be Dynamic or Thomas-Cook inspired packages to the customer at point of them they require it.
Customers can re-open a dynamic or fixed package and continue editing or viewing it before booking the package.
Customers can re-open a dynamic package and continue editing it before booking the package.
Goal: Create a central memory system where users can manage, re-open, and share all their packages effortlessly.
Itinerary Model & Rules
To prevent the AI from generating impossible trips, I defined a block-based itinerary system. Each day is made of blocks tagged as:
Arrival
First block of Day 1
Departure
Last block on last day
Accomodation
Check in/Check out into hotels and Overnight Stays
Activity
Attractions, Tours, and Experiences
Food
Restaurants and Meal Stops
Transport
Inter-city or Inter-country travel
Image 1
Dynamic Package Itinerary Example
This model keeps the UX clean while giving engineering a clear schema for AI-generated itineraries.
Visual Design System
Following Thomas Cook’s brand, I created a modern, AI-native UI layer:
Color Rationale
Voyage Blue (#0056B3)
Symbolizes trust, reliability, and forward momentum. As the primary brand color, it anchors the interface with a confident, travel-ready tone and reinforces Thomas Cook’s legacy of dependable service.
Sunshine Gold (#FFC107)
Represents energy, warmth, and moments of delight. Used as the accent color, it highlights key actions, important states, and interactive elements—guiding users naturally without overwhelming the interface.
Cloud Mist (#D9E6F2)
A soft, airy neutral that provides clarity and contrast throughout the UI. Ideal for panels, backgrounds, and layered surfaces, helping the conversational and itinerary elements feel clean, modern, and easy to scan.
Type Rationale
Aa
Bold Weight for Section Titles
Poppins
Regular Weight for Body and Chat text to keep conversations readable
Layout - Desktop first Responsive Grid
Left: Navigation & quick access to My Chats / Packages
Center: Chat conversation as the primary canvas
Right: Contextual pane for packages, filters, and itinerary editing
From Flows to UI (Process)
1
Identity & Design Brief - I formalized the problem space, goals, and competitive landscape (Ask Layla, PickYourTrail, Airbnb Experiences).
2
Workflow Mapping - Defined the six key workflows and their triggers, states, and success criteria (Destination exploration, Dynamic packages, Fixed packages, My Chats, Packages tab, Sharing).
3
Itinerary Rules & Edge Cases - Codified what the AI can and cannot change inside an itinerary.
4
Wireframes & Conversation Flows - Designed message sequences for how the chatbot collects preferences, announces when it has “all key details”, and then directly shows relevant packages instead of over-explaining.
5
High-Fidelity UI & Handoff - Built componentized Figma layouts for chat, panes, filters, and package cards, aligning with Atirath’s engineering stack (React).
Learnings & Impact
01
Designing for Conversations, Not Screens
Shifting from page-based navigation to dialogue-driven flows meant designing every question, response, and confirmation as a UX artifact, not just the UI container around it.
02
Guardrails Make AI Useful
The itinerary rules and non-deletable structural blocks turned “AI magic” into reliable, bookable trips instead of random suggestions.
03
Continuity = Trust
“My Chats” and “Packages” turned AI from a one-off gimmick into a long-term planning companion. Users always feel like the system “remembers” them.
04
Bridging UX & Front-End
Because the flows and visual system were designed with React implementation in mind, the project now serves as a template for future AI products within Thomas Cook and Atirath’s broader Tacy ecosystem.
Transforming Thomas Cook’s static tour packages into an AI-powered, conversational travel planning experience.
Tacy AI Ecosystem
Role
Lead UX/UI Designer & Product Strategist
Company
Atirath Technologies x Thomas Cook India
Tools
Figma
FigJam
REACT
Notion
Miro
Launched: January 2025
Atirath Website
Starting Point: Legacy Experience
Thomas Cook’s digital presence was built around static, pre-defined packages and traditional booking flows. Users had to jump between:
01
Marketing pages for destination research
02
PDF Itineraries for package details
03
Separate forms or call centers for modifications and bookings
There was no single, continuous experience where a traveler could explore, customize, and book a trip without starting over each time they changed their mind.
Project Overview
TravBridge (internally, part of the Tacy AI Ecosystem) is a Generative AI–powered travel assistant that helps Thomas Cook customers plan vacations from scratch or using curated packages, all inside a conversational interface.
The system contains
An AI chatbot for destination discovery and trip planning
A dynamic package builder that updates itineraries in real time
Fixed Thomas Cook–inspired packages for fast, low-effort bookings
Support for revisiting old chats, saved packages, and sharing itineraries with co-travellers
Outcomes at a glance
Designed an end-to-end chat-first UX for travel research, package creation, and booking.
Defined 5+ core workflows (explore destinations, dynamic packages, fixed packages, My Chats, Packages, sharing).
Created a visual design system aligned with Thomas Cook (blue / yellow / white) and a desktop-first product UI ready for engineering handoff.
Established itinerary rules & constraints so AI-generated packages remain realistic and bookable.
The Challenge
How do you let users:
Talk to a chatbot like a travel agent
Customize trips endlessly
Still keep the structure, reliability, and compliance of Thomas Cook packages tours
Key Challenge 1
Designing a conversation model that feels natural but still collects all required trip data (destinations, dates, duration, travellers, budget, preferences).
Key Challenge 2
Balancing full customization (dynamic packages) with zero-effort options (Thomas Cook–inspired fixed packages).
Key Challenge 3
Ensuring continuity so users can pause and come back later without losing any planning work (My Chats, Packages tab).
Key Challenge 4
Encoding itinerary logic (arrival, departure, hotels, activities, transport) so the AI never creates impossible schedules.
Design Goals
01
Gen Z-friendly, AI-native UX
Make the chatbot the primary interface, not an add-on.
02
Single Continuos Journey
From Vacation inspiration to travel package customization to booking the travel package without context loss.
03
Two planning modes
Effortless curated packages vs. deep customization, both within the same UI.
04
Safe, realistic itineraries
Encoded rules so the AI respects travel constraints and package standards.
05
Scalable Design system
Flows and UI components that Thomas Cook can reuse for future AI features (visa rules, forex, etc.).
Ecosystem Map: Core Functionalities
From the design brief, I grouped TravBridge into five core pillars:
AI -Powered Destination Exploration
Dynamic Package Creation (Fully Customizable)
Thomas-Cook inspired Fixed Packages
My Chats – Revisit & Resume Conversations
Packages Hub – Saved & Booked Packages + Sharing
Each pillar became a dedicated workflow with its own triggers, UI states, and chatbot tone.
Key Workflows
01
Explore Destinations via Chat
Users start a new conversation from the AI Chat Helper / “New Chat” entry points.
The chatbot asks about budget, travel style, and interests (adventure, nightlife, off-beat, nature).
It returns short, scannable cards for each destination: best season, key attractions, visa conditions, and typical budgets.
Follow-up questions refine recommendations instead of sending users to new pages.
Goal: Replace scattered research with one focused, conversational flow.
02
Create Dynamic & Personalized Packages
This is the heart of the Tacy / TravBridge experience.
The chatbot offers the customer whether they want to create a fully personalized package or explore Thomas Cook packages first.
For a dynamic package, A persistent right-hand pane opens with an editable package shell where the customers can edit key properties of the package and personalize it according to their liking.
The itinerary generated by AI is bound to rules: Arrival and Departure blocks, Accommodation per night, Activities, Food, and Transport types.
Goal: Give travelers full control to shape their itinerary while ensuring the AI maintains structure, accuracy, and bookability.
03
Thomas-Cook Inspired Fixed Packages
The chatbot collects various key preferences from the customer and TravBridge surfaces pre-designed Thomas Cook packages with fixed itineraries.
Each packages shows various details to the customer like Duration, key cities, trip theme, inclusions, exclusions, visa conditions, and terms & conditions.
Customers cannot modify these itineraries but they can simply book the package with in a few simple clicks in the AI ecosystem’s interface.
Goal: Offer effortless, ready-to-book packages for users who prefer fast decisions over detailed customization.
04
My Chats: Continuity in planning
Since Trip planning happens over days or weeks, Every conversation is stored in a My Chats tab with filters inspired by Airbnb’s filter patterns (destination, date, status).
Reopening a chat restores the full message history and all the packages that have been discussed in the thread.
The chatbot can resurface older packages or suggest new package options if preferences have changed since that chat.
Goal: Let users return to any trip plan at any time without losing context, progress, or previously generated packages.
05
Packages Hub: Saved, Booked, and Shared
A dedicated Packages tab becomes the memory of the ecosystem.
The Packages hub shows the saved/booked packages whether it may be Dynamic or Thomas-Cook inspired packages to the customer at point of them they require it.
Customers can re-open a dynamic or fixed package and continue editing or viewing it before booking the package.
Customers can re-open a dynamic package and continue editing it before booking the package.
Goal: Create a central memory system where users can manage, re-open, and share all their packages effortlessly.
Itinerary Model & Rules
To prevent the AI from generating impossible trips, I defined a block-based itinerary system. Each day is made of blocks tagged as:
Arrival
First block of Day 1
Departure
Last block on last day
Accomodation
Check in/Check out into hotels and Overnight Stays
Activity
Attractions, Tours, and Experiences
Food
Restaurants and Meal Stops
Transport
Inter-city or Inter-country travel
Image 1
Dynamic Package Itinerary Example
This model keeps the UX clean while giving engineering a clear schema for AI-generated itineraries.
Visual Design System
Following Thomas Cook’s brand, I created a modern, AI-native UI layer:
Color Rationale
Voyage Blue (#0056B3)
Symbolizes trust, reliability, and forward momentum. As the primary brand color, it anchors the interface with a confident, travel-ready tone and reinforces Thomas Cook’s legacy of dependable service.
Sunshine Gold (#FFC107)
Represents energy, warmth, and moments of delight. Used as the accent color, it highlights key actions, important states, and interactive elements—guiding users naturally without overwhelming the interface.
Cloud Mist (#D9E6F2)
A soft, airy neutral that provides clarity and contrast throughout the UI. Ideal for panels, backgrounds, and layered surfaces, helping the conversational and itinerary elements feel clean, modern, and easy to scan.
Type Rationale
Aa
Bold Weight for Section Titles
Poppins
Regular Weight for Body and Chat text to keep conversations readable
Layout - Desktop first Responsive Grid
Left: Navigation & quick access to My Chats / Packages
Center: Chat conversation as the primary canvas
Right: Contextual pane for packages, filters, and itinerary editing
From Flows to UI (Process)
1
Identity & Design Brief - I formalized the problem space, goals, and competitive landscape (Ask Layla, PickYourTrail, Airbnb Experiences).
2
Workflow Mapping - Defined the six key workflows and their triggers, states, and success criteria (Destination exploration, Dynamic packages, Fixed packages, My Chats, Packages tab, Sharing).
3
Itinerary Rules & Edge Cases - Codified what the AI can and cannot change inside an itinerary.
4
Wireframes & Conversation Flows - Designed message sequences for how the chatbot collects preferences, announces when it has “all key details”, and then directly shows relevant packages instead of over-explaining.
5
High-Fidelity UI & Handoff - Built componentized Figma layouts for chat, panes, filters, and package cards, aligning with Atirath’s engineering stack (React).
Learnings & Impact
01
Designing for Conversations, Not Screens
Shifting from page-based navigation to dialogue-driven flows meant designing every question, response, and confirmation as a UX artifact, not just the UI container around it.
02
Guardrails Make AI Useful
The itinerary rules and non-deletable structural blocks turned “AI magic” into reliable, bookable trips instead of random suggestions.
03
Continuity = Trust
“My Chats” and “Packages” turned AI from a one-off gimmick into a long-term planning companion. Users always feel like the system “remembers” them.
04
Bridging UX & Front-End
Because the flows and visual system were designed with React implementation in mind, the project now serves as a template for future AI products within Thomas Cook and Atirath’s broader Tacy ecosystem.