01
We manage change together with you.
Before we propose anything, we go where the work actually happens. We train and strengthen the team’s knowledge, and we frame the diagnosis in numbers.
Atypical Design Studio
We design strategy and interactions.
Working groups are a business model that allows you to get different perspectives from experts in the market.
49+
Organisations
839
Workshop hours
6
Markets
2010
Designing since
Is this for me?
Worried that change is too big a nut to crack?
02
We assemble the team
In working groups we prepare diverse approaches to your problem. Then we connect you with partners whose solution you choose.
03
We stay until the end
We don’t disappear after handing over documentation. We prepare the team for change, support migration, and make sure every capability of the shipped tool is intuitive and used to the full.
Where are you?
Different companies. Different problems.
Different first steps.
Every engagement is built to succeed.
01
People development and human skills
People development programmes for mid-size and large companies that know technology without competence is a cost, not an investment.
BUR · KFS · DigComp 2.2 · Training
Development programmes02
Fast deployments for SMBs
Delivery from day one. Licensed stacks, safe environments. QuickWins for SMBs means outcomes in days and weeks, not quarters — without a giant programme or a long contract.
Automation · Integrations · First AI steps
Minimal sensible pilot (PoC)03
ML / IoT in operations and industry
You have a shop floor, machines and data you don’t use? Integrate the knowledge and unlock the potential of what you already have.
Computer vision · Edge computing · Prediction · AI in industry
Data & ML programme04
Claude + MCP — introduction to AI
Two-day programme: LLM fundamentals, privacy and licensing, Claude surfaces (Chat, Cowork, Design, Code), Projects, Connectors/MCP, Skills — and responsible use in the team.
Claude · MCP · Skills · Enterprise
Claude + MCP training05
Architecture with your own IT and infrastructure
Architecture design and solution choices for companies with internal teams and infrastructure. We join as advisors and fit your design and delivery process.
Architecture · Due diligence · Vendor selection
Technical architecture06
Augmentation.
You don’t have to do it in the dark.
Augmentation supports your own work — for owners, managers and specialists who want to work smarter without giving up control. Tools that strengthen how you think, not replace it.
Prompting · AI workflows · Personal work systems
Augmenting your own work03 · Case study
Change starts with people
We don’t define projects through a tools-first lens. We look at what changes for the client — and what should stay unchanged.
Vision
Form
Intelligence
Process
From problem to concrete transformation
We do not start with a single forced path — first we understand your situation, then we show sensible options within your budget.
-
Report your problem
-
Proposal of first steps
-
Internal training
-
We assemble a working group
-
You receive up to 3 solutions and approaches for your problem within your budget
Training enrollment
Local AI Vision from scratch — model, tracking, events
Book your seat at training
Open registration: Claude + MCP training
Announcement
Alint in Berlin. Polish Tech Night.
Blog
IT contract safety for the buyer
Check whether your contract is not a machine built for long-term milking of the client.
Blog
You ask who will deploy AI for us. The real question: does your company know how it actually works?
From the WordPress era to GPT, we keep running into the same pattern.
Announcement
Alint working group and project kick-off
Announcement
Training on AI in brand building: enrollment closes.
Announcement
Alint working group and project kick-off
Blog
Poland is short an estimated 50,000–150,000 ICT specialists. Even if you hire one, your company may not be ready.
For eighteen months you may pay senior invoices while getting a translator. Here is how to shorten that runway and grow architects inside the organisation.
Announcement
Alint — tender opens under Olsztyn Science and Technology Park
Blog
The dead end of “we will just train our own model”
When a narrow ML model is a bullseye, what a language calculator is, why training your own LLM on contracts is a blind alley, and three questions before you spend the budget.
Announcement
Helverth — working group launch
Blog
The migration took eight months. A year later, the headaches were the same.
New stack, new UI — and the same friction on pricing, logistics, and data. A “what you are importing” checklist, four moves before the RFP, and one numeric question before you sign.
Announcement
2026 enrollment opens — September. Join a working group.
Blog
Technical debt is not a technical problem
Technical debt is a ledger of past trade-offs. When nobody is left to fight fires, the culprit is often a chain of earlier shortcuts — not headcount alone.
Blog
Load more
Solutions
Solution views we are building
Products and working-group projects — click through to explore.
Voice CRM · CarPlay · field
Alint
The invisible CRM for field teams — a voice note after a meeting becomes client knowledge.
See Alint →
CMS · NGO · reporting
CMS_4_NGO
Green Morning NGO CMS — updates, disclosure duties and project history from day one.
Open the preview →
Measurement · production · ROI
Helverth
Measurement and classification in production — savings estimates instead of manual counting.
See Helverth →
PoC · LLM · contextual reading
Cursor Adapt
Contextual cursor — every word becomes a knowledge node. Optimistic UI and pre-fetch mask AI latency.
Open the PoC →Case studies
Organisational change stories
New brand strategy & rebrand · 2020
Grodno S.A.
Changing how the company thinks about itself and its assets.
2020. Poland’s largest electrical wholesaler. Thirty years on the market, eighty‑eight branches. The category was shifting — renewables, photovoltaics, EV charging, new partner types at the table.
Together with Grodno we designed a new mark and guided them through the change — because visual change without human change is just a new PDF. Seven hundred people had to feel the new direction was theirs, not imposed from outside.
Thanks to the team: Piotr Łysakowski — marketing and strategy; Bartek Mitura — PM; Kacper Rzosiński — designer; Karolina Pawlina and Łukasz Fedorowicz for collaboration and counsel.
We did not build the website. The momentum that followed was entirely theirs.
We did not build the website. The momentum that followed was entirely theirs.
Revenue growth is the work of a whole organisation. The rebrand was one element of that story — not its cause. Data: GPW: GRN.
1.19B
Revenue 2021/22
+92%
Growth since 2019/20
+28%
E‑commerce YoY
2×
Industry awards (year 1)
B2B2C platform delivery + sales prediction
Euromaster Nordic
Changing B2B2C operations.
A new system of work for sales.
Delivered with Frontkom. Euromaster (Michelin Group) was serving B2B customers, dealers and franchisees through a collection of disconnected systems. At peak season, the team was spending thousands of hours re-entering data between platforms.
We delivered a unified B2B2C platform combining sales, invoicing and document flow — integrated with external tooling. The Laravel backend supports complex discount structures and re-invoicing logic.
Delivered 2018–2021. Total project cost recovered by end of 2022.
Delivered by Frontkom with UX by Atypical Design Studio.
200,000€
savings in two years
+41%
revenue vs previous year
-38%
manual work
Back‑office SaaS — audit & MVP redesign
All4Cloud / APN Promise
Managing thousands of licences — redesigned in five weeks.
All4Cloud is a Microsoft licence back-office serving administrators, resellers and tenants. Built over years by an internal team alone. At scale, it began to buckle: navigation three levels deep, expiring contracts buried, campaign conflicts, licence errors distributed across the interface with no apparent logic.
Twenty-five working days. Audit, UserTasks workshop, Blueprint Model, TreeJack testing via Optimal Workshop, MVP in Figma. Output: a Critical Actions panel, simplified approval workflow, granular role permissions, dark mode, and redesigned data tables with sticky headers, horizontal scroll, and components drawn from Material and IBM Carbon.
Kacper Rzosiński × Frontkom, 2019. Based on post-launch user surveys.
30%
time saved on licence management
+40%
user satisfaction
25 days
from audit to MVP
Does this sound familiar?
For sixteen years we have watched capable people stand before their boards and explain
with considerable discomfort, why the new system failed to move the needle — and why nobody uses it.
You have been in that room.
Whatever your context above — these are the patterns that keep showing up in the same rooms.
Scenario 01
A proposal that protects the vendor, not you
The less defined your objective, the more you pay for someone else’s peace of mind about their own uncertainty.
Your subcontractor protected themselves from every risk they could see. No one at the table was protecting you.
Scenario 02
The AI that never leaves the demo
Nine months of work. Impressive demonstration. Three months later, the system collects dust — because it was never designed around how people actually work, and fails in roughly thirty per cent of real cases.
AI without Gemba is simply a fashionable solution searching for a problem to justify it. AI is not always the answer.
Scenario 03
A rebrand that stops at the logo
In eighty per cent of cases, a genuine brand change is an internal one: it must reshape sales operations, processes and the way people behave day to day. It must serve B2B, B2C and the team simultaneously.
A new colour palette and a generative‑AI wordmark will not change how an organisation thinks — nor give it a direction worth thinking about.
Scenario 04
A product that excludes fifteen per cent of your market
Shipped. Budget spent. Customers calling with complaints. Accessibility was never built in. Retrofitting it costs three times as much — plus the ongoing operational cost while customers wait.
WCAG is not a compliance checkbox. It is a quality threshold. And it is considerably cheaper as a first-day decision than a last-day apology.
Scenario 05
A platform that can take an order
It worked. The customer could complete a purchase in thirty seconds. Technically, a success. Nobody asked why the drop-off rate keeps rising, or why average order value has been falling quietly for two years.
Commerce is not a button that says “Buy now.” It is everything that happens in the customer’s mind before they consider clicking it.
Scenario 06
Replacing platforms on an annual cycle
The migration consumed eight months and an entire two-year budget. A year later, the new platform carried the same problems as the old one — because nothing in the organisation around it had changed.
The constraint is rarely the platform. More often, it is the absence of a clear plan for what the platform was supposed to enable.
05 — How we work
Six elements. One direction.
We do not work from a catalogue. Every engagement begins with a single question: what must this organisation be prepared for in three years — and what, precisely, stands in the way?
We start with Gemba. Always.
Gemba — Japanese for “the real place.” We go where work actually happens, not where reports about it are compiled. We diagnose in numbers, not in adjectives.
The problems we find at Gemba are almost never the ones the client described in the first meeting.
Gemba — go to the floor
User interviews, process observation, shadowing.
Numbers-first diagnosis
Time, cost, risk. We define the problem measurably before proposing anything whatsoever.
Minimal intervention
We select tools to fit the problem. No single methodology serves every situation.
We stay to the end
We support implementation and return with a report: did the change produce the outcome it was designed to produce?
Vision is the first thing to align.
We design strategy and communication for organizations entering a new chapter—new market, new offer, new partner. We answer: what must the company be ready for in three years? How will customers, employees and partners react?
We don’t do greenwashing. We audit reality first—then build a step-by-step transition you can defend.
Strategy workshop
An extended sprint with leadership. Intensive work on direction and decisions.
Qualitative research
Focus groups + 1:1 interviews about vision, constraints and real blockers.
Comms & reach audit
Channels, message effectiveness, brand monitoring and awareness signals.
Training cycle (optional)
Quarterly program for leadership: change management, tech horizon, expert panels.
Engage people in change
Soft rollout of participation; optional internal transformation leads.
Recommendations & rollout plan
Internal + external comms plan, and clear guidelines for agencies.
Pilot
Implementation + monitoring for 3–9 months; for ESG companies—extended financing advisory.
ESG without greenwashing. Only numbers.
ESG reporting is increasingly a regulatory obligation—and often a requirement to enter supply chains. We build ESG matrices from scratch: baseline audit, KPI mapping to ESRS/GRI, and a plan you can defend to an auditor.
Companies that start ESG with comms instead of data—pay twice.
Baseline audit
Data collection: emissions, energy, supply chain, policies, governance—no polishing.
Reporting framework
ESRS (CSRD), GRI, UN SDGs—fit to your obligations and industry.
KPI matrix
Measurable indicators for E/S/G with data sources and cadence.
Gaps & remediation plan
Where data is missing or below threshold—realistic timeline and owners.
Report & documentation
Ready for publishing and financing conversations; audit-friendly trail.
Operational costs are design decisions.
OPEX grows not from bad intent but from processes designed for another scale, team or market. We map flows and identify where the organization pays for chaos. Recommendations always come with numbers.
The most expensive process is the one nobody questions because “we’ve always done it this way”.
Process mapping
Value stream mapping—where waiting, handoffs and waste live.
OPEX analysis
We break costs by concrete actions—not departments.
Bottlenecks
Where does flow stop? Which decisions are made too high or too low?
Recommendations with a bill
Effort/impact prioritization with estimated change cost and return.
Implementation & monitoring
We track results after 30/60/90 days and adjust based on data.
Tools don’t work without people.
Implementations fail not because tools are bad, but because the organization isn’t ready. We audit digital capability by teams and roles—not job titles—and design development paths for real gaps.
Digitalization without capability is buying gym equipment nobody uses.
Digital capability audit
Gap analysis with DigComp 2.2 and industry context.
Roles & gaps map
Required level per role and what blocks transformation.
Learning paths
Microlearning, shadowing, practical workshops—fit to the gap.
Pilot program
Measure capability change before/after; scale what works.
Internal digital leaders
Develop ambassadors so the organization can grow without us.
Technology serves the problem. Not the other way around.
We keep a proven toolkit and know when to use what. ML background (2018), VUI/CarPlay/chatbots/computer vision before they became fashionable. Currently building Alint—voice CRM assistant for field sales on CarPlay.
We’re often 2–3 years ahead of clients. That’s not an accident.
UX & Interaction Design
Design systems, information architecture, research, prototyping.
VUI — Voice User Interfaces
CarPlay, mobile voice, conversational assistants.
AI implementation
Readiness audit, automation, LLM integrations, AI agents.
WCAG 2.1 / 2.2
Accessibility audits, inclusive design, public-sector statements.
06 — Projects
Ours and clients’
Some projects are active. Some are concepts that never launched—each left a method we still use.
Alint
Voice CRM assistant on Apple CarPlay for field sales. A 3‑minute voice session instead of evening form filling. Supported by WA‑MA incubator grant.
HELVERTH
A sports/streetwear brand for teens. Bold geometric language and authenticity-driven communication.
Fuzzy
Electronic jazz festival identity and comms system for a project that never launched—method survived.
Cocole
Premium tea packaging UX and identity—where physical experience meets digital communication.
F1 Telemetry Platform
F1 telemetry visualization for fans and analysts—turning technical data into narrative without engineering background.
In collaboration with Aleksander Zawalich
Jbet
Racing betting system—UX designed responsibly in a space that typically designs for addiction.
08 — Before you write
Not for everyone.
Quite literally.
We work with organisations that sense their category is moving — and wish to be prepared before the rest of the market notices.
- You are entering a new segment or market, and neither the existing brand nor the existing processes will carry you there.
- You can commit to a budget decision within thirty days — we do not prepare proposals for filing cabinets.
- You want a partner who will say no when the direction is wrong.
- You measure outcomes — you care about P&L, not about appearances.
- You have a Budget Cap for implementation investment — without a defined upper limit, the project does not start.
- You can commit at least PLN 20,000 to the workshop cycle — without it, direction and scope cannot be built honestly.
- You are not running “layoff first” programmes — we do not design processes with redundancies as the primary objective.
- You are not doing greenwashing — we do not take on ESG work whose goal is communication without data.
- You do not want dark patterns — we do not design manipulation into procurement, sales, or onboarding.
Twenty minutes. Three questions. We will know whether we can help.
connect@atypical.pl Book a conversation