When I joined Onfly as Head of Design, a stakeholder diagnosis revealed six critical pain points: inconsistent experience, non-existent processes, low design maturity, lack of product vision, poor cross-functional collaboration, and difficulty meeting deadlines.
I treated it as a project — structured plan, prioritized fronts, execution in cycles.
Context
Onfly is Latin America's largest B2B travel tech company. The Product Design team plays a central role: turning business complexity into simple, efficient experiences for users, operating at the intersection of technology, product, and customer experience.
When I took over as Head of Design in March 2025, the team had only two Product Designers — not enough to cover the complexity of the products or the quality level the business demanded. One of the first decisions was to grow the team intentionally: hire the right profiles, distribute them across the highest-impact fronts, and create a DesignOps function that didn't exist — to ensure governance and consistency across design, product, and technology. The team had talent. What it lacked was scale, structure, and architecture.
My first move was to make no structural decision before listening. I spent the first month in conversations with stakeholders and key people across the company to understand pain points, perceptions, and expectations around the UX practice and the product experience.
The Diagnosis
Those conversations mapped six core pain points:
Poor and inconsistent user experience across the app and platform — no visual or interaction standards across surfaces.
No journey mapping or personalization — the team was delivering screens, not connected experiences.
Low collaboration between Design, Marketing, Product, and Engineering — design operated in silos, with no cross-functional alignment rituals.
Lack of product vision and low design maturity — designers had no clarity on the strategic impact of their work on the business.
Lack of transparency, organization, and process definition — no documentation model, no structured ceremonies, no delivery visibility for other areas.
Low delivery quality and difficulty meeting deadlines — a symptom of the above, not the root cause.
A seventh gap emerged from the structural analysis: the absence of a DesignOps function. There was no one responsible for the operational bridge between Design, Product, and Technology — process governance, delivery consistency, file organization, and traceability were all spread across product designers with no clear ownership.
With this map in hand, I organized the action plan across seven fronts: roles and responsibilities, culture and organizational structure, training and development, research and strategy, processes and work methodology, strong design culture, and planning and governance.
The Approach
I organized execution around the same three pillars I use to lead: Direction, Process, and People.
Direction — making the chapter visible to the business
I built an OKR framework for the Design Chapter mapped directly to the company's Tech & Product objectives (NPS 54→62, scaling operations without scaling cost, delivery quality and speed), identifying where design had direct ownership, where it had a supporting role, and where there was no territory to claim.
In parallel, I initiated journey and flow mapping and structured alignment with the Product Marketing team (PMM) to connect design discoveries to Go-to-Market strategy. This produced something that hadn't existed before: a clear answer to the question "what is Design doing and why does it matter to the company right now?"
Process — building systems that reduce dependency on individuals
Formalized UX process. I built the Product Design process at Onfly from scratch — defining upstream and downstream, discovery stages, problem definition, prototyping, validation, and delivery, along with a delivery checklist and documentation model. The process made the chapter's work trackable, predictable, and communicable to product and engineering stakeholders.
Creation of the DesignOps function. One of the most important structural decisions was establishing a dedicated DesignOps role — a function that didn't previously exist at the company. The role was created to serve as the operational bridge between Design, Product, and Technology: responsible for Design System governance, delivery process consistency, file organization and traceability, and the interface with engineering teams across tribes. Without this function, governance would have remained diluted across product designers, making any process improvement fragile and dependent on individual goodwill.
Rituals and ceremonies. With the DesignOps function active, I implemented the team's ritual agenda: alignment ceremonies, a file organization model, and documentation processes to ensure standardization, visibility, and scalability across tribes.
Design System diagnosis and strategic restructuring. Before making any changes, we conducted a formal DS maturity diagnosis — mapping gaps between Figma and production, inconsistencies across tribes, and the absence of guiding principles. From the diagnosis, we defined four principles to guide all system decisions: Natural (experience adapted to device, native patterns as default, 8pt scale), Intentional (clear hierarchy, minimal visual noise, visible CTAs), Inclusive (WCAG AA accessibility as default, valid contrast, keyboard navigation), and Unmistakable (colors, typography, tone, and icons forming a recognizable Onfly signature).
With the principles in place, we structured the first DS Sprint — with an explicit RACI, defined deliverables (principles consolidation, visual territory exploration, system naming, lockup, and foundations setup in Figma), and governance distributed across tribes rather than a single central point.
NPS as a strategic signal. I restructured the NPS program — from tool selection (Survicate) to a weekly synthesis format that generated actionable insights for leadership in under 30 minutes. Result: baseline NPS of ~47 in Q4/25. Q1/26 closed at 53 (+6 points). Q2/26 ranged between 52 and 59.
AI-first workflow. I positioned the Design team as a driver of Onfly's AI strategy — not just a consumer of it. The logic: design is the contact point between people and technology, which puts us in the ideal position to explore how AI can improve real experiences. I structured the adoption of Claude Code for real-time DS adherence validation, Figma Make for handoff compression, Lovable for functional prototype and proof-of-concept generation, and Claude Design for interface exploration and generation directly within the creation workflow. I also defined the Harness Design vision: repositioning the Design System as context infrastructure for LLMs, with intent-based documentation per component, semantic tokens, and autonomous decision guides.
People — building the foundation for autonomy
Career plan and ladder. I built the chapter's career structure from scratch: hard skills (interaction design, visual design, product vision), soft skills (communication, ownership, resolutiveness), and Onfly competencies defined per level. The ladder covers the full progression — from Designer I to Director — with specific responsibilities per level in each dimension, turning growth into a concrete path rather than a vague promise.
Leveling and salary review. I applied formal leveling to the team based on the defined ladder, corrected existing asymmetries, and aligned salary parameters to market benchmarks — creating a fair and transparent foundation for growth conversations.
Tour of Duty and IDP. I introduced the Tour of Duty as a bilateral expectations contract between designer and leadership — a clear agreement on individual development goals and expected contributions over 12 to 18-month cycles. Each designer gained clarity of direction, purpose, and active ownership of their own growth. The Individual Development Plan (IDP) operationalized that contract into concrete actions following the 70/20/10 model: 70% practical learning through real challenges, 20% social learning through feedback and mentorship, 10% formal learning through courses and reading.
Performance framework. I conducted the team's first structured performance evaluation cycle, producing calibrated positioning for all members, promotion justifications for three roles, and structured monthly feedback for the full team.
Onboarding and capability building. I structured the onboarding process for new members with business knowledge scalability to accelerate the ramp-up curve. I facilitated internal workshops on communication and negotiation, product strategy, and AI tool adoption.
Results
Area | Before | After |
|---|---|---|
Diagnosis | No formal pain point mapping | 6 pain points mapped with stakeholders, 7 action fronts |
Career plan | No ladder, no levels, no defined hard/soft skills | Full ladder from Designer I to Director with responsibilities per level |
Leveling & salary | No framework, existing asymmetries | Leveling applied, salary parameters aligned to market |
Development | No IDP, no Tour of Duty, no structured onboarding | Bilateral TOD, IDP 70/20/10, formalized onboarding |
DesignOps function | Non-existent | Role created with ownership of DS governance, processes, and Product & Tech interface |
Design process | No upstream/downstream, no documentation, no ceremonies | End-to-end process, rituals, checklist, documentation model |
Design System | No diagnosis, no principles, centralized governance | 4 principles defined, Sprint 1 delivered, tribe-based governance |
NPS | Manual, 3–4h/week, no strategic output | <30min/week, weekly leadership brief, +6 pts Q1/26 |
Performance | No formal framework | Evaluation cycle, promotions for 3 roles, structured feedback |
AI workflow | No structured adoption | Claude Code, Figma Make, Lovable, Claude Design, Harness Design vision |
Chapter direction | No OKR visibility | Chapter mapped to Q3 company OKRs with named contributions |
What this taught me
The stakeholder diagnosis was the most important decision I made. Arriving with answers before listening would have been the most common mistake — and the most costly. The six pain points mapped in that first month guided absolutely everything that followed.
The second lesson was about sequence. With so many simultaneous fronts, the discipline was in prioritizing what reduced the most friction in the team's day-to-day before building what raised the strategic bar. You can't ask for excellence from people who don't know what's expected of them or which direction they're heading. The career plan came before the performance review — not the other way around.
The third: a leader who becomes the single point of failure has built a fragile team, not a strong one. Every structural decision here — transparent ladder, bilateral Tour of Duty, governance by proximity, documented processes — was oriented toward the same outcome: a chapter capable of operating, evolving, and delivering without me needing to be in every room.
