SYSTEMS
Built a production-ready design system in an engineer-first startup
With no precedent, no tooling and no initial buy-in. What started as a rejected pitch became a company culture switch with an adoption program in development.
I was hired as a UI Designer on a greenfield product: a Marketing Desktop CRM ∂that was to become the company's second pillar of their product suite.
Within the first weeks, it became clear there was no existing product definition: no UX, no feature scope, no patterns: just a rough direction and a proof-of-concept deadline.
During onboarding, I went throught the company’s existing first CRM product. What I found wasn’t a rough design system: it wasn’t even a component library.
With a new clean product starting from zero, this felt like a rare opportunity to build things the right way from the start at a system level. That’s where the idea of a design system originated.
Partial backing by VP
↓ ↓ ↓
Company-wide rollout
The design system was not part of the projects initial scope.
I pitched it early, but was rejected by CEO;the priority was shipping features, and design infrastructure was not seen as a potential improvement.
However, the proposal did resonate with the VP of Engineering, who was acting as frontend coordinator and JavaFX specialist; this gave me technical backing and a clear signal - the system could exist as long as it helped delivery and didn’t slow the product down -
Japan timezone and only written english
→ Result: Full async methods pushed to the test. Building the system incrementally alongside design work made friction areas smaller and easier to deal with.
The product was starting from zero, but it wasn't in a vacuum. Senapt already had a in-production CRM product. During onboarding, and first small tasks, I got to interact with it and understand how design and development worked. What I found was a warning sign:
No design system, no unified filing systems, libraries, styles or components. This was the outcome of years of designers coming and going and accumulating legacy and perferences one on top of each others.
> Scattered components and styles
> No use of variables
> No Styleguide or library
> No single source of truth
What we inherited:
The strongest validation came from daily collaborators, not executives.
→ Result: We built the product specs Through design and iteration. Each step was validated.
Partnership, early validation, and open communication.
Short-term fixes weren't sustainable.
Not everything worked; what did became the base.
Documentation replaced constant alignment and allowed us to jump over language and time gaps.
Building created the momentum that discussion couldn't.
This whole project proved the value
of a systematic approach.
A pain-point surfaced during validations and was owned end-to-end, from discovery to delivery.
Social Media managers were constantly pulled into handling customer complaints arriving via Facebook messages. These were manually handled and incresingly blurred the line between marketing and support.
The idea was to use the conversation itself to identify support cases and connect them directly to the CRM.
Facebook message → ID request → CRM validation → Support workflow
When a customer sends a Facebook message, an automated reply prompts them to provide their contract number if the issue relates to support. Once detected, the ID is validated against the CRM and the customer is automatically located.
The Marketing Management platform delivered measurable impact across three dimensions: Product, Organization, and Business. This later reflected in CEO buy-in and investors-facing use.
This project didn’t just reach production-ready. It changed how I work.
I joined the company as a mid-level UI designer, with no prior product or leadership experience.
For the first time, I found out firsthand what it really means to build while defining. Scoping, designing, shipping, validating, revisiting, and restructuring all happened at once, with no clean lines between them.
I had to navigate the high level of uncertainty most startups are known for. Design, product, and system concerns were all active at the same time, constantly influencing one another.
At times, it was chaotic and stressful. Early decisions were continuously tested while foundations were still being laid; priorities shifted, assumptions broke, and not everything held.
Processes were sometimes messy. Acting as a one-person team – covering UI, UX, systems, product decisions, and parts of project management – forced me to step into responsibility quickly, with no buffer between decisions and consequences.
Looking back, the most valuable outcome wasn’t a feature, metrics, or a set of polished screens. It was learning how to keep moving without fully knowing the road ahead, and growing into responsibility through doing – not waiting for a title or label to arrive first.
Approved look & feel - First screen design
User journey walkthrough
Card hover interactions
Social inbox UX - Interaction patterns
Social inbox UI - Interaction design
Social inbox - Handoff documentation
Brand identity excerpt
Campaigns dashboard UI
Create campaign - Review summary
Energy sector lead details
Login flow animation