3
Platforms design simultanoudly
200k+
Active users served by Sky 2.0 chatbot
100%
On-schedule design delivery across sprints
~40%
Faster research-to-wireframe with AI workflows
I own the design layer end-to-end — from requirement gathering through to wireframes, high-fidelity prototypes, design systems, validation, and developer handoff — partnering closely with Product Managers, engineers, BA teams, and business stakeholders to turn requirements, API constraints, and business objectives into feasible, shippable experiences.
Role Context
Core Methodology
End-to-End Product Design, the Double Diamond Way.
Every revamp — whether a single flow or a whole vertical — moves through the same four phases: discover the real problem, define the right one to solve, develop a range of solutions, and deliver a validated, dev-ready design.
Challenge
Four verticals had grown independently, with inconsistent IA, ad-hoc design decisions, and no shared process — making it hard to predict delivery timelines or maintain quality across teams.
My Approach
Introduced Double Diamond as the working framework for every initiative — grounding decisions in discovery evidence (heuristic audits, flow data, stakeholder input) rather than opinion, and closing every project with validation before handoff.
Outcome
A repeatable process adopted across 4 concurrent verticals, giving PMs and engineering predictable delivery checkpoints and giving design a consistent seat earlier in the roadmap conversation.
DISCOVER
Problem Discovery
Heuristic evaluation, funnel/flow audits, and stakeholder input to surface friction across Flights, Hotels, Holidays, and Europamundo.
DEFINE
Information Architecture
Reframing findings into clear problem statements, IA maps, and user flows that scope what needs to be designed.
Discover
Problem Discovery
Heuristic evaluation, funnel/flow audits, and stakeholder input to surface friction across Flights, Hotels, Holidays, and Europamundo.
DEVELOP
Wireframes & Prototypes
Exploring options through wireframes and high-fidelity prototypes, built on a shared Flutter-compatible design system.
Discover
Problem Discovery
Heuristic evaluation, funnel/flow audits, and stakeholder input to surface friction across Flights, Hotels, Holidays, and Europamundo.
DELIVER
Validation & Handoff
Validating with stakeholders and research signals, then packaging specs for a clean, 100% on-schedule developer handoff.
Live - 200K+ Users
Sky 2.0 — AI-Assisted Chatbot, Shipped.
Designed and shipped Sky 2.0, an AI-assisted chatbot experience within the mobile app, powering self-service journeys for 200K+ active users — from first artboard through to production handoff.
Challenge
Most post-booking queries (cancellations, status checks, refund status) routed to human agents, creating slow response times and rising support load as the user base grew.
My Approach
Designed a conversational UI with quick-reply chips to minimize free-text input (which the AI model handled least reliably), and treated fallback/low-confidence states as first-class screens — not an afterthought — so failed self-service degraded gracefully into human handoff.
Outcome
Shipped to production and now supports 200K+ active users, shifting a meaningful share of routine queries away from human agents.
Component 01
Conversation Interface
Message bubbles, bot vs. user differentiation, typing indicators, and self-service intent chips.
Component 02
Quick Replies & CTAs
AI-assisted flows for booking queries, cancellations, and support — reducing dependence on human agents.
Component 03
Edge States & Fallbacks
Graceful handling of low-confidence responses, timeouts, and handoff to human support.
Live on the Akbar Travels app, supporting self-service for 200K+ active users.
Status
Live in Production
UX Across Every Surface.
Ongoing UX changes across mobile app, web, and M-Site — spanning Neo-CFAR, Air Legit, Cabs integration, and Holiday Flight — with 100% on-schedule design delivery maintained across every sprint.
Challenge
Four unrelated initiatives — refunds, trust, cross-sell, and flow-merging — needed to ship in parallel across app, web, and M-Site without design becoming the bottleneck for any engineering pod.
My Approach
Prioritized by business/user urgency each sprint, reused shared design-system components wherever possible to cut production time, and kept a tight feedback loop with engineering on API constraints before high-fidelity work started — avoiding late-stage rework.
Outcome
100% on-schedule design delivery maintained across every sprint, with all four initiatives shipped to production.
100% on-schedule design delivery maintained across sprints — even while running four concurrent workstreams across three platforms.
Delivery
Neo-CFAR
Redesigned cancellation and fare-adjustment/refund experience — clarifying fare rules and reducing support escalations.
Air Legit
Trust and verification layer for flight bookings — surfacing legitimacy signals at the right moment in the funnel.
Cabs Integration
New cross-sell vertical designed and integrated into existing booking journeys across app and web.
Holiday Flight
Combined flight + holiday package flow — unifying two previously separate booking journeys.
In Progress
Redesigning Flight & Hotel — Less Noise, Clearer Choices.
Completed high-fidelity product flows for Flight, Hotel, Holidays, and Europamundo, while actively redesigning the Flight and Hotel experiences to improve information architecture and reduce cognitive load at decision-heavy moments.
Challenge
Search-results screens for Flight and Hotel had accumulated filters, badges, and fare details over time, creating high cognitive load right at the decision moment — the highest-risk drop-off point in the funnel.
My Approach
Audited every element on the results screen against a single question — "does this help the user decide right now?" — then applied progressive disclosure, pushing secondary detail (fare rules, amenities) behind a tap instead of showing it upfront.
Outcome
High-fidelity flows completed for all 4 verticals; Flight and Hotel redesigns currently in build, targeting a clearer primary decision path and reduced scroll depth.
100% on-schedule design delivery maintained across sprints — even while running four concurrent workstreams across three platforms.
Delivery
Focus 1
Simplified Information Architecture
Restructuring search, filters, and results so travellers reach a decision with fewer competing signals on screen.
Focus 2
Reduced Cognitive Load
Progressive disclosure of fare rules, add-ons, and hotel amenities — surfacing detail only when it's relevant.
Focus 3
High-Fidelity Flows
End-to-end flows completed for Flight, Hotel, Holidays, and Europamundo — from search through confirmation.
Focus 4
Bus Module Integration
Adapting the design language for Bus module product, within shared design-system constraints.
In Progress
Sharpening Navigation Through Heuristic Evaluation.
Collaborated UX improvements for the Akbar Travels and Tickut websites, using heuristic evaluation and UX research insights to strengthen navigation, information hierarchy, and the overall digital experience.
Challenge
Akbar Travels site had a formal usability review in place — navigation and IA issues were being reported anecdotally, with no structured way to prioritize fixes. Tickut website was to be developed from scratch.
My Approach
Ran a full heuristic evaluation against Nielsen's usability principles on both sites, cross-referenced findings with existing UX research, and ranked issues by user impact vs. effort to build a prioritized fix list stakeholders could act on.
Outcome
Navigation and information hierarchy restructured on key entry points across both sites, based on evidence rather than opinion.
Method 1
Heuristic Evaluation
Systematic review of both sites against usability heuristics — surfacing navigation and hierarchy gaps.
Method 2
UX Research Synthesis
Combining evaluation findings with research insights to prioritise the highest-impact fixes.
Method 3
Navigation & IA Fixes
Restructured navigation and information hierarchy across key entry points on both sites.
In Progress
Europe Fixed Departure — Extending the Holidays Experience
Created three tailored Home-page variations for the Europe Fixed Departure product, while keeping the rest of the journey aligned with the existing Holidays booking flow.
Challenge
Europe Fixed Departure needed a distinct entry point within the Holidays experience, without creating an entirely separate booking journey or adding unnecessary complexity.
My Approach
Explored and designed three Home-page variations specifically for Europe Fixed Departure, adapting the entry experience to communicate the product clearly while maintaining consistency with the existing Holidays flow and interaction patterns.
Outcome
A focused Europe Fixed Departure experience that gives the product its own presence on the Home page while leveraging the established Holidays journey for a consistent and familiar booking experience.
Reframed dense, itinerary-heavy content into a scannable structure — reducing pre-booking questions and improving confidence in the Holidays booking journey.
Content Design
On Going Process
AI-Assisted Design — ~40% Faster Iteration.
Using AI-assisted design and research workflows to accelerate research-to-wireframe activities, improving personal design iteration speed by approximately 40% — without cutting corners on rigour.
Challenge
Manual research synthesis and first-pass wireframing were the slowest part of every project's Discover-to-Develop transition, limiting how many concepts I could explore before committing to a direction.
My Approach
Used AI tools to cluster raw research and heuristic notes into structured problem statements, and to generate rough first-pass wireframes from IA inputs — then refined and validated everything by hand, treating AI output as a faster starting draft, not a final answer.
Outcome
Roughly 40% faster research-to-wireframe cycles, with the time saved reinvested into exploring more directions per problem rather than shipping fewer iterations.
Speed gains were reinvested into more exploration per problem — not fewer design cycles — keeping quality and rigour intact while moving faster from research to a testable wireframe.
Outcome
Step 01
Research Synthesis
AI-assisted clustering of research notes and heuristic findings into structured problem statements.
Step 02
Rapid Wireframing
Faster first-pass wireframes generated from IA and flow inputs, then refined by hand.
Step 03
Faster Iteration Loops
Shorter cycles between concept, critique, and revision — roughly 40% quicker than manual-only workflows.
Designing at the intersection of scale and speed.
What this work demonstrates
1. End-to-end UI Design of an AI-assisted feature — Sky 2.0 — serving 200K+ active users daily
2. Running the Double Diamond process consistently across four product verticals at once
3. Maintaining 100% on-schedule delivery while leading UX across app, web, and M-Site
4. Using heuristic evaluation and research to drive measurable navigation and IA improvements
6. Using heuristic evaluation and research to drive measurable navigation and IA improvements
7. Turning AI-assisted workflows into a real ~40% speed gain without losing design rigour
Where I'm growing next
1. Deepening quantitative validation — tying redesigns like Flight/Hotel directly to funnel metrics
2. Extending Double Diamond discovery earlier into roadmap and feature ideation conversations
3. Formalising the AI-assisted workflow into a repeatable practice the wider team can adopt
4. Expanding the Flutter design system's coverage into edge-case and empty states
5. Expanding the Flutter design system's coverage into edge-case and empty states