Overview
BairesDev is Latin America's largest technology outsourcing company and a leader in nearshore delivery, with 3,000+ engineers across 36 countries serving clients including Google, Rolls-Royce, Johnson & Johnson, Pinterest, and ViacomCBS.
For a company of that model, the website is not marketing collateral — it is the primary acquisition surface on both sides of the business. It has to convince a CTO that BairesDev can deliver mission-critical engineering, and simultaneously convince a top-percentile engineer that BairesDev is a place worth working. Both audiences arrive suspicious of exactly the thing the site is claiming.
I joined as UX Design Specialist and led the redesign of the institutional ecosystem: 120+ core pages, 80+ landing pages, and the Daily Bundle blog publishing new content every day, all across multiple languages. I ran discovery, mentored a team of three designers, built the design system, and worked alongside engineering — including hands-on front-end contribution — to keep implementation faithful to design intent.
Business Context
Outsourcing is a trust purchase in a low-trust category. Enterprise buyers have heard the nearshore pitch before, often after a bad experience with a vendor who over-promised. The site's job is to make credibility legible before a sales conversation ever happens.
Three business realities shaped the work:
Two funnels, one property, opposite proof requirements. Enterprise clients need evidence of delivery capacity, technical depth, and reliability. Engineers need evidence of interesting work, real career progression, and rigorous hiring standards. Both funnels run through the same domain, and the content that reassures one can undercut the other — talent-marketing language reads as fluff to a CTO, while enterprise sales language reads as corporate to an engineer.
Scale made inconsistency inevitable without a system. 200+ pages across multiple languages, plus daily blog publishing, meant content was being created faster than any review process could police it. Consistency had to come from reusable structure, not oversight.
Traffic at 200k monthly visitors made design decisions measurable — and consequential. At that volume there was enough behavioral data to prioritize on evidence rather than opinion, and enough at stake that a poorly-performing template compounds across hundreds of pages.
Why This Was Hard
This wasn't a segmentation problem solved by separate sections. Positioning that maximizes enterprise confidence — emphasizing scale, process, and managed delivery — is precisely what makes a strong engineer suspect the work is commoditized. The redesign had to hold both propositions credibly on one property.
With daily blog output, any design solution requiring per-page design attention would fail within a week. Templates and patterns had to absorb unpredictable content without degrading.
Text expansion, localized navigation, and content variance across languages meant layouts had to be designed for variability rather than for a canonical language and adapted afterwards.
UX methodology was not yet an established practice across the company's digital properties. The redesign's scope depended on securing C-level buy-in first — which meant building an argument in analytics and interview evidence before any design direction could be proposed.
Mentoring three designers through discovery, prototyping, and delivery ran concurrently with my own design output and front-end contribution.
My Role and Ownership
I led the redesign of the institutional ecosystem across discovery, design system, team leadership, and implementation:
Led discovery end to end — 40+ stakeholder and user interviews, plus analytics review across 200k+ monthly visitors — establishing where the ecosystem was losing each audience and why.
Secured C-level buy-in for UX methodology, converting research findings into an executive-level argument that expanded UX adoption beyond this project into the company's broader digital practice.
Mentored and directed a team of three designers through discovery, prototyping, and delivery, structuring the work so the team grew while shipping.
Built the design system that maintained consistency across 200+ multilingual pages and a daily-publishing blog.
Produced and validated wireframes and prototypes from low to high fidelity, testing directionally with stakeholders before committing to build.
Contributed front-end code alongside developers in agile sprints — closing the gap between design intent and shipped implementation directly rather than through specification handoff.
Approach and Key Decisions
Lead with evidence to earn scope.
I sequenced discovery deliberately: interviews and analytics first, design proposals second. With UX not yet established as a company practice, the research wasn't only diagnostic — it was the instrument that secured executive support for a redesign broad enough to actually fix the problem rather than restyle it.
Design distinct pathways, not distinct sites.
Rather than splitting client and talent experiences into disconnected properties, I designed differentiated entry points and journey logic on shared infrastructure. Each audience gets proof calibrated to its own skepticism, while the ecosystem stays coherent, maintainable, and consistent in brand.
Treat the design system as publishing infrastructure.
With daily content and 200+ pages, the system's purpose was throughput. Templates, components, and content patterns were built so marketing and content teams could publish at cadence without design becoming a bottleneck or consistency degrading over time.
Prioritize by analytics, argue by interview.
Quantitative data showed where users dropped, hesitated, and left; the 40+ interviews explained why. Combining both meant every prioritization decision had a measurable justification and a human explanation — the pairing that made design arguments persuasive to executives who think in metrics.
Design for language variance from the start.
Multilingual constraints were treated as primary layout inputs rather than a localization phase, so components tolerated text expansion and content variation by construction.
Stay in the code.
Contributing to front-end implementation meant design intent survived the build. On a project this large, fidelity loss compounds — every template shipped slightly wrong multiplies across the pages that use it.
Design System Impact
The system was the mechanism that made scale manageable:
Coverage across 200+ multilingual pages, giving both audiences a consistent experience regardless of which team or language produced a given page.
Publishing-ready templates for the Daily Bundle blog, absorbing new content daily without per-post design work.
Localization-tolerant components, built for text expansion and content variance rather than assuming a canonical language.
Landing page patterns reused across 80+ campaign pages, making new campaigns fast to launch and consistent by default.
A shared design–development vocabulary, reinforced by my own front-end contribution, which reduced translation loss between design and implementation.
Extended UX adoption, as the system gave non-design teams a way to make design-consistent decisions without design review.
Outcomes
Measurable conversion improvement on client inquiries from redesigned landing pages — design work tied directly to the company's commercial acquisition funnel.
+35% average session duration within six months, indicating visitors were finding and engaging with content rather than bouncing off it.
−27% bounce rate across top-performing pages.
+18% increase in job applications from high-value candidates in the first quarter post-launch, confirming the talent-side pathway worked alongside the client-side one.
Daily publishing cadence maintained with no UX or performance bottleneck — the design system operating as intended under sustained load.
UX methodology adopted more broadly across the company's digital properties, extending the project's impact past the redesign itself.
Key Learnings
Analytics get you the meeting; interviews get you the mandate. Executives engaged with the numbers, but the interview findings were what made the problems feel real enough to fund properly. Neither alone would have secured the scope.
When two audiences conflict, design the pathways — don't average the message. The instinct to find messaging that works for everyone produces messaging that convinces no one. Differentiated journeys on shared infrastructure served both without diluting either.
At publishing scale, the design system is the product. Anything requiring per-page design attention fails immediately against daily output. The system's job was throughput, not aesthetic consistency.
Multilingual is an architecture decision, not a translation step. Designing for text variance from the beginning avoided the far more expensive alternative of retrofitting layouts that only worked in one language.
Writing code protects the design. Front-end contribution wasn't scope creep — on a 200-page ecosystem it was the most reliable way to ensure what shipped matched what was designed.
Closing Statement
BairesDev was a lesson in designing for scale and for conflict simultaneously: an ecosystem large enough that only systems could hold it together, serving two audiences whose requirements pulled in opposite directions. The redesign moved both funnels — and the more durable outcome was that UX stopped being a project the company ran once and became a method it kept using.