Category
UI/UX • Front End Development
Brand
CodeDXB
Year
2026
Timeline
~2 wks
Project Brief
The client needed a conversion-focused conference website for CodeDXB the UAE's flagship AI and technology summit targeting enterprise decision-makers, C-suite executives, and regional tech leaders.
The brief was straightforward but the pressure wasn't: build a platform that could handle multi-audience registration flows (sponsors, exhibitors, visitors, media), communicate authority in a competitive regional market, and go live fast. No existing design system. No brand UI history. Just a logo, a vision, and a deadline.
My role: Sole UI/UX Designer and Frontend Developer.
Constraint: End-to-end delivery: design, build, QA, and deploy.
The Core Problem
The UAE tech conference market is saturated with high-budget competitors like GITEX, Expand North Star, that have years of brand equity and dedicated web teams behind them. A new entrant like CodeDXB couldn't compete on scale yet, but it could compete on clarity, credibility, and conversion.
The core problem: How do you make an enterprise audience trust a brand-new conference enough to register, sponsor, or exhibit?
The previous digital presence didn't answer that question. Additionally, it was built from a Wix website so I have to migrate the contents. The old website has no clear hierarchy of who the platform was speaking to, no distinct value proposition for each audience type (visitor vs. sponsor vs. exhibitor), and no design language that signaled premium positioning.
Baseline Audit, Before Redesign
| Metric | Baseline |
|---|---|
| Avg. session duration | 1m 12s |
| Bounce rate | 71% |
| CTA click-through rate | 4.2% |
| Registration form completion | 18% |
| Mobile engagement rate | 31% |
Baseline data pulled from Google Analytics and Search Console
Competitive Research
I audited four competitors across local and international markets to identify gaps and define positioning.

The Solution
Before opening Figma, I needed to understand who was actually using this platform and what they needed to be able to do. I ran a WhatsApp thread and a Google Meet session with the internal stakeholders and account managers who were closest to the client, since direct exhibitor access wasn't available at this stage.
The questions I focused on:
I collated the answers in a Figma whiteboard by grouping pain points by audience type (visitor, sponsor, exhibitor) and mapping what each group needed, in what order, to move toward a conversion action.
From there I built a user flow diagram before touching any UI. Only once that flow was validated did I move into high-fidelity design.
Frontend execution ran in parallel with final design revisions, with QA and deployment on Vercel closing the project out.


Brand Guidelines
The CodeDXB identity anchored around the tension between code and enterprise where technical credibility meets the executive audience. For the visual system, I collaborated with a graphic designer to collate my visions and stakeholder's visions and this is what I used all throughout the project:
Color: Near-black base (#0F0C05) for primary text, electric-blue accent (#0000EA) for CTAs and electric pink (#E501A0) for highlights only.
Typography: Clean Objectivity hierarchy, large editorial headings, tight body copy, generous whitespace. Designed to feel like a publication, not a brochure.
Tone: Direct, confident, industry-peer. Not salesy. The copy treats the reader like they already belong in the room.

Tools I Used
I used Figma for brand assets, icon design, and visual references. Adobe Illustrator handled logo refinements and vector assets. For copy drafting and content structuring, I leaned on Claude to move fast without sacrificing quality. The entire frontend was built in Windsurf, with GitHub for version control and repo management, and Vercel for deployment and hosting.
The Outcome
The platform launched on time and it's live right now, actively running for the 2027 event cycle. Real organisations like ATEX International, Mutant, Tech-Z Magazine, The Agenda, have their logos publicly on the site. That matters more than it might sound. For a brand-new conference with no history, having companies with real reputations choose to associate publicly with your platform is the credibility signal the brief was actually asking for. It showed up before the post-launch metrics even came in.
When the numbers did come in, session duration more than doubled, bounce rate down 38%, CTA click-through up 178%, mobile engagement more than doubled, registration completion from 16% to 38%. What stood out to me was that the baseline data had basically predicted where those improvements needed to happen. A 74% bounce rate and a 19% CTA click-through on the old platform wasn't a traffic problem. People were arriving. They just couldn't figure out what the platform was for or where they belonged in it. That single read from the data is what drove the audience segmentation decision more than anything else I did on this project.
Post-launch Results (vs. baseline)
| Metric | Baseline | After Launch | Change |
|---|---|---|---|
| Avg. session duration | 1m 12s | 2m 48s | +133% |
| Bounce rate | 71% | 44% | −38% |
| CTA click-through rate | 4.2% | 11.7% | +178% |
| Registration form completion | 18% | 41% | +128% |
| Mobile engagement rate | 31% | 67% | +116% |
Baseline data pulled from Google Analytics and Search Console
What I Took From This
The audience segmentation like visitors, sponsors, and exhibitors have different needs. But the previous platform treated them all the same and the bounce rate was the proof that it wasn't working. Once each audience had their own path and their own answer to the question they came with, the registration numbers followed.
The speed of delivery was also a real outcome, not just a nice timeline stat. Getting the platform live fast meant the client could spend their energy on what actually builds a conference for speakers, partnerships, and filling seats. They weren't sitting around waiting for a website while the event date got closer.
If I could go back, I would have run a website test on GITEX and Expand North Star to give real users an actual task to complete on each competitor site and watching where they got stuck or sailed through. That behavioral data would have made the IA decisions even more defensible than my visual audit alone. And a five-second test on the homepage hero before committing to the full build direction would have told me quickly whether the premium positioning was actually landing on first impression or just feeling right to me.



































