Loading...
Luděk Černocký/cernocky.com01 / 36
From Code to Company · 2nd April 2026From Code to Company
From Code to Company
Designer's approach
This is me!
Design Leader. Mentor. Human.
Building design teams, shipping products, and guiding talent for 20 years.
2021–present
VP Product Design•Emplifi
2012–2021
Head of Design•Socialbakers
2006–2012
Head of Product & Design•eStránky.cz
2004–2006
Freelancer
Company context
SaaS · B2B/B2CCX platform · 10+ products98% desktop600+ employees
emplifi.design
14 membersProduct Design · UI & Ops · Research
Design?
“
Design is solving problems
— Paul Rand
“Design is solving problems
Design is solving problems
with a set of constraints
— Mike Monteiro
Mental models
Microinteractions
Customer journey
Wireframes
Typography
User flows
Discovery
Design systems
User-testing
Usability
Affordance
Design tokens
Double Diamond
Accessibility
Jobs To Be Done
Experiments
Layout
Design critique
Personas
Prototyping
Information architecture
Cognitive load
User research
A/B testing
Design sprints
Google these. Or ask AI. Either way, faster than me :)
Design is not decoration
See the problem before you code itTest ideas before they become expensiveMake products easier to understand and use
A designer who can't explain why something exists is just decorating.
The Builders
Roles are blurring
PM
Design
Engineering
The best builders don’t wait for handoffs. They think across product, design and code.
The Product Builder
01Understand the problem
02Pick what to test
03Build the smallest thing
04Learn from real users
Before you write code,
answer these 3 questions
Who is this for?What problem does it really solve?How will you know if it works?
The goal is not to slow down building. The goal is to avoid building the wrong thing too confidently.
A rough prototype and a few honest reactions can save weeks of engineering time.
Experiments
We have 5000 files in Figma
We do not have 5000 products or features
Wait, so what are we "designing"?
Conceptsto test if an idea deserves to exist
Testsprototypes built to learn, not to impress
MVPsthe bare minimum to prove value
Visionsa direction, not a commitment
Branded Demos
Design System
Presentations
Onboarding
Hype
Website(s)
From 0 to hero. Care in 3 years.
This was Care in April 2022
Care prototype in May 2022
...and we built this in 3 weeks
Testing with our first client
They tested it internally...
...and with another client
Validating the idea
01The idea was a hypothesis
02A prototype made it testable
03Real users made it credible
04Revenue made it real
“
Experimenting is way cheaper than building s**t nobody needs
— me
Same approach today for Agentic AI
Building AI directly in code
We call it AIUI — AI takes predefined inputs and generates outputs.
Rapid prototyping directly with clients
Lessons & tips
Strategy
When everyone can build,
strategy becomes the advantage
Execution is easier than ever. Choosing the right problem is still hard.
Lack of market need remains one of the most common reasons startups fail (CB Insights).
Business
You can't build the right thing
if you don't understand the business
Know who pays and why they stayUnderstand the business model before proposing solutionsMeasure your work in outcomes, not outputs
The best engineers and designers don't just ship features — they understand how those features drive revenue.
Learning speed
The fastest team is not the one that ships first, but the one that learns first
A prototype that fails in two days saves you two months of development time.
Iteration
The power of discarding ideas
Don't fall in love with your first draftExplore multiple variants before committingKill ideas early when they are cheap
If you haven't thrown away at least 3 variations, you probably haven't explored enough.
Judgement
When output is cheap,
judgement becomes expensive
Tools can generate options. Judgement is choosing the right one for the right reason.
AI can give you 100 options in a minute. It still takes a human to pick the right one.
Visual quality
In a world of similar tools,
trust is visible
Quality signals careClarity signals competenceConsistency builds credibility
People perceive aesthetically pleasing design as more usable (Aesthetic-Usability Effect).
Usability
If people don't get it,
the idea doesn't work
Make the next step obviousRemove friction before adding featuresTest with users before writing too much code
Testing with just 5 users uncovers 85% of usability problems (Nielsen Norman Group).
Test before you build.*
Learn fast.
Build smart.
* In the new world of AI, it might be faster to build stuff, so this might not be entirely true anymore.
Thank you
Now it's your turn
Ask me anything :)
