Was this a Beabstracto project?

No. Thiago Andrade was permanently employed as a UX/UI designer at izz.ai in Munich from October 2021 to February 2023. The case describes that work; Beabstracto takes method and mindset from it, not client relationships or products.

Which products or clients are named?

None. izz.ai's client projects and products remain confidential. The case describes the kind of work — B2B SaaS interfaces, dashboards, AI-driven product concepts — without naming individual products or showing screens.

Are there verified figures?

No. No cleared figures on usage, revenue or time saved exist from this period. We do not quote numbers we cannot back up, so the outcome is described qualitatively.

What does SaaS and AI product design have to do with specialist systems in banking and industry?

The situation is the same: a new feature or module needs to reach users quickly, but requirements are immature and stakeholders lack a shared picture. Flows, prototypes and an implementation-ready handover solve that at an asset management company, a software vendor or in a control room just as they do for a SaaS product.

How did the collaboration with development work?

Direct and continuous. Thiago worked with developers and product owners on flows, behaviour and states, and handed over high-fidelity designs with handoff files and UI specifications. The aim was that nobody has to guess behaviour in short delivery cycles.

Which roles is this case relevant for?

For heads of product, product management and project leads at software vendors who want a feature — often with AI — accompanied from concept to handover. And for recruiters and intermediaries looking for a UX/UI designer or product designer who covers discovery, requirements clarification and implementation-ready design in one person.

SaaS & AI · B2B products · 2021–2023

Turning vague product ideas into usable SaaS interfaces — as the only designer.

At izz.ai in Munich, Thiago Andrade worked from October 2021 to February 2023 as an employed UX/UI designer, shaping SaaS solutions for SMEs and enterprise clients across Europe. Immature requirements and early product ideas became user flows, high-fidelity prototypes and implementation-ready designs. This case shows how that works without a large design team — and what Beabstracto takes from it for its clients.

Book an intro call How we went about it

Client: izz.ai (employer, Munich)
Industry: Digital products, SaaS and AI initiatives for SMEs and enterprise clients across Europe
Role: UX/UI Designer, permanent employment — sole designer
Period: October 2021 – February 2023
Systems: B2B SaaS interfaces, SaaS dashboards, AI-driven product concepts
Team: Founders, stakeholders, product owners and developers; design in one pair of hands

In short

izz.ai is a Munich-based company that built digital products, SaaS solutions and AI initiatives for a range of clients. Thiago Andrade was its only designer there from October 2021 to February 2023, employed as a UX/UI designer. The problem: product decisions had to be made before requirements were mature — in a fast, client-facing environment without an established product structure. His job was to translate business ideas, client needs and product concepts into usable B2B SaaS interfaces, interactive prototypes and implementation-ready designs. The result: vague ideas became concrete interfaces, stakeholders could decide on a prototype, and developers received handoff files and UI specifications. No verified figures exist, and none are estimated.

For readers in product management or a business department this is familiar territory: a new feature — often with AI — needs to reach the customer quickly, but nobody has yet thought it through in flows and screens. That translation work is exactly what Beabstracto brings today to projects with software vendors, banks and industrial companies.

Starting point

The starting point was a small, fast, client-facing product environment in which design could not wait for finished requirements.

izz.ai worked with different clients on digital products, SaaS solutions and AI-driven initiatives — for small and medium-sized businesses as well as enterprise clients across Europe.

Thiago joined in October 2021 as UX/UI designer and was, in that role, the central and usually only designer. There was no large design team and no mature product structure to lean on. Design decisions, user flows and interface behaviour sat in one pair of hands — from the first idea to the handover to development.

The users of the resulting interfaces were business users at the clients: people who work with SaaS dashboards and B2B interfaces and who expect an AI feature, above all, to be understandable and useful in their daily work. On the other side stood founders, stakeholders, product owners and developers who needed a product direction they could plan, build and sell against.

Before this work, product ideas lived mostly in conversations, presentations and technical reasoning. What was at stake: a client does not buy a concept they cannot picture, and a developer cannot build anything robust from an idea whose behaviour nobody has defined. Design had to do both — fast enough for the business, structured enough for implementation.

The problem

The problem was that product decisions came before the requirements — and nobody was closing the gap between business idea and usable interface.

Requirements not mature
Projects started with a goal, some business logic and a lot of open questions. Which steps a user actually goes through, which data they see and what happens when something fails was often undefined by the time the first interface was needed.
No design team behind him
As the only designer there was no review round, no research colleague and no mature product structure. Every decision about flows, components and states had to be made, justified and documented — alone.
AI capability without a user's view
In AI-driven initiatives the starting point was what the technology could do. For end users it was unclear what the system does, what they enter, what comes back and how far they can trust the result. Without that translation, an AI feature stays a bullet point on a slide.
Stakeholders without a shared picture
Founders, clients and product owners talked about the same product and meant different things. Without something tangible, there was no way to align internally or to show clients convincingly where the product was heading.
Speed versus product logic
Delivery cycles were short. The risk: interfaces that look finished quickly but have no consistent logic — and developers who have to guess behaviour because the handover has gaps.

How we went about it

Thiago's approach: ask the right questions early, make ambiguity visible quickly in flows and prototypes, and carry every decision through to the handover to development. This was work in permanent employment — the method is the same one Beabstracto uses today as an external partner.

1
Product discovery with stakeholders
Every project began in direct conversation with founders, clients, product owners and developers: what is the goal, what are the constraints, what business logic sits behind it, what outcome is expected? Because time was short, the questions had to be early and precise. The outcome was a shared understanding of purpose, audience and boundaries — the basis on which requirements could be clarified at all.
2
Translating requirements into user flows
What had been said became user flows and wireframes. This step exposed gaps: missing states, unclear decisions, steps nobody had considered. Product requirements, user flows and interface behaviour were sharpened together with the stakeholders instead of being discovered during development.
3
High-fidelity prototype as the basis for decisions
For alignment, what counted was what people could touch. Thiago built interactive high-fidelity prototypes that let stakeholders understand the product direction, test assumptions and present ideas to clients. The prototype was thus both a validation and a sales instrument — before the first line of code.
4
Making AI capabilities usable
In AI-driven product concepts the task was to translate emerging technical ideas into interfaces and workflows that end users understand and find valuable. What does the user enter, what does the system return, how is the result presented — these questions were answered in product concepts and prototypes, not in specifications alone.
5
Implementation-ready handover
Finally, designs, flows and specifications were prepared for development: high-fidelity UI, handoff files and UI specifications for behaviour and states. The aim was for developers to build in short delivery cycles without guessing behaviour — and for the product logic to survive the transition from design to code.
6
Repeating across projects and domains
This cycle ran in parallel across several client projects in different domains. Switching between contexts forced a way of working that does not rely on prior knowledge of each field but on fast orientation: who uses this, what has to come out at the end, which decision is blocking right now?

Outcome

The outcome was that vague product ideas became concrete SaaS interfaces that stakeholders could align on and developers could build. No verified figures on usage or revenue exist from this period, so none are given.

Aspect	Before	After
Product idea	Lived in conversations, slides and technical reasoning	Tangible as user flow, wireframe and high-fidelity interface
Requirements	Incomplete; gaps surfaced only during development	Gaps surfaced in flows and prototypes and were closed with stakeholders
Stakeholder alignment	Debate over terms, everyone with their own picture	Decisions made on an interactive prototype
AI feature	Technical possibility without a user's perspective	Understandable workflow with clear input and output
Handover to development	Room for interpretation, follow-up questions in short cycles	Handoff files and UI specifications for behaviour and states

From vague idea to concrete interface
Across several client projects Thiago delivered design work in a fast-moving environment and helped turn fuzzy product ideas into concrete B2B SaaS interfaces and dashboards. The interfaces balanced clarity for business users, speed of delivery and technical feasibility.
Alignment through prototypes rather than meetings
Prototypes and implementation-ready designs gave stakeholders something to decide on. Product direction was settled before implementation, not during it.
Autonomy as a designer
Without a large design team, a robust independence emerged: responsibility for flows, interface behaviour and handover in one pair of hands, in direct collaboration with founders, product owners and developers.
What this case proves
That design can be owned end-to-end in ambiguous environments, that speed need not cost product logic, and that business ideas can be translated into usable SaaS experiences — even when AI is involved.

What we learned

01
Asking early beats correcting late
When requirements are not mature, the most important design contribution is not the screen but the question: who uses this, what has to come out, what happens when it fails? The earlier these questions are asked, the cheaper the answer. At Beabstracto this is now the as-is analysis at the start of every engagement — at a software vendor just as much as in a business department at a bank or an industrial company introducing a new system.
02
The prototype is the functional spec everyone reads
Few people read a requirements document; everyone understands a clickable prototype — founders, product owners, the business department, development. At izz.ai the prototype decided and sold. In implementation projects it plays the same role: the business department sees before go-live what is coming, and development builds against something concrete.
03
AI only becomes a product when users know what it does
Technical capabilities only become valuable when input, output and reliability are understandable to the user. That holds for a young SaaS product as much as for an AI module in a banking or industrial system: without that translation the feature goes unused, however good the model.
04
Speed needs structure, not less of it
Short delivery cycles tempt you to drop specification. The opposite was needed: pin down flows, states and behaviour so that developers do not guess. That is why implementation-ready handover is part of every prototype at Beabstracto — not an extra after the design.
05
Being effective without a large team
As the only designer you cover roles that are usually spread across people: discovery, concept, design, handover. For clients this means Beabstracto does not need a well-oiled design team on the other side to have an effect, and fits into existing product and project teams instead of demanding a structure of its own.

What we delivered

User flows
Wireframes
High-fidelity UI designs
Interactive prototypes
SaaS dashboards
Product concepts for AI-driven initiatives
Design handoff files
UI specifications
Stakeholder presentation material

Frequently asked questions

Was this a Beabstracto project?: No. Thiago Andrade was permanently employed as a UX/UI designer at izz.ai in Munich from October 2021 to February 2023. The case describes that work; Beabstracto takes method and mindset from it, not client relationships or products.
Which products or clients are named?: None. izz.ai's client projects and products remain confidential. The case describes the kind of work — B2B SaaS interfaces, dashboards, AI-driven product concepts — without naming individual products or showing screens.
Are there verified figures?: No. No cleared figures on usage, revenue or time saved exist from this period. We do not quote numbers we cannot back up, so the outcome is described qualitatively.
What does SaaS and AI product design have to do with specialist systems in banking and industry?: The situation is the same: a new feature or module needs to reach users quickly, but requirements are immature and stakeholders lack a shared picture. Flows, prototypes and an implementation-ready handover solve that at an asset management company, a software vendor or in a control room just as they do for a SaaS product.
How did the collaboration with development work?: Direct and continuous. Thiago worked with developers and product owners on flows, behaviour and states, and handed over high-fidelity designs with handoff files and UI specifications. The aim was that nobody has to guess behaviour in short delivery cycles.
Which roles is this case relevant for?: For heads of product, product management and project leads at software vendors who want a feature — often with AI — accompanied from concept to handover. And for recruiters and intermediaries looking for a UX/UI designer or product designer who covers discovery, requirements clarification and implementation-ready design in one person.