Designing system logic. Clear paths for users.

I combine qualitative and quantitative methods. First, I analyze product metrics — funnels, drop-off points, task completion time — and then enrich them with qualitative insights from customer development, interviews, and usability testing.

Based on this, I frame hypotheses using JTBD and Job Stories so they are directly tied to real user needs.

I evaluate hypothesis success through predefined metrics: A/B testing, changes in key KPIs, conversion rates, task completion time, error rates, and CSI/CSAT.

Finally, I compare results against the baseline and decide whether to scale or iterate the solution.

In uncertain situations, I rely on a mix of product thinking, qualitative insights, and fast hypothesis validation.

If data is missing or contradictory, I first outline several possible hypotheses using JTBD and user experience patterns, then refine them through quick qualitative methods like interviews, prototyping, and usability testing.

I then choose the option with the lowest risk for both the user and the business — usually the one that is easiest and fastest to validate.

After implementation, I always define metrics and validate outcomes through analytics or A/B testing to confirm or adjust the decision.

In one B2B project, I used JTBD while redesigning a CRM ticket workflow.

Interviews revealed that users didn’t really care about changing statuses — what mattered was quickly understanding the current customer state without losing context.

This shifted the design approach: instead of adding more features, we simplified the user flow, consolidated key actions into a single screen, and restructured the customer card.

As a result, task completion time decreased and the number of user errors dropped.

In a B2B CRM project, I worked on the ticket processing module.

Analytics and user feedback showed that the main workflow took too long and was error-prone due to a fragmented interface.

I redesigned the ticket structure, unified key actions into a contextual flow, and simplified the user journey using CJM and CustDev insights.

After release, ticket processing time dropped by ~25–30%, errors decreased, and support requests for this flow were reduced.

I use AI as an accelerator for research, design, and content production.

ChatGPT helps me structure hypotheses, JTBD, Job Stories, UX copy, and interview synthesis.

Codex OpenAI, AI Studio, and VSCode are used to build interactive prototypes and React / TypeScript / Tailwind-based concepts.

Gemini and NotebookLM help generate UX alternatives, explore flows, and prototype ideas quickly.

Google Flow Pro (Nana Banana) is used for visual concepts, while Kimi helps prepare presentations.

In my projects, I designed AI-driven scenarios for B2B and ERP systems, mainly focused on assistance, content generation, and simplifying complex workflows through AI-assisted UX.

My role included identifying user scenarios where AI actually reduces cognitive load.

I also designed AI flows and interface states, defined JTBD and Job Stories for AI features, and built and tested interaction prototypes.

Additionally, I worked on edge cases, error handling, and user trust in AI outputs.

I used CustDev, CJM, JTBD, scenario modeling, rapid prototyping, and usability testing.

First, I clarify the reasons behind the disagreement — whether it’s business constraints, metrics, risk, or different problem framing.

Then I align it with user data and product goals to identify where the mismatch actually is — in the solution or in the problem definition.

I either defend the solution with evidence from research, CJM, metrics, and testing, or propose alternatives and compromises.

If data is insufficient, I quickly validate assumptions through prototyping or testing and return with evidence.

I’ve mostly worked with backstage and support processes, as they have the biggest impact on stability in B2B and ERP systems.

The main challenge is synchronizing internal processes across teams — when users expect instant actions, but the system requires multiple service steps or approvals.

In such cases, I focus on identifying delays and information gaps, aligning SLAs and responsibilities, and simplifying dependencies.

I also improve system feedback and make internal states more transparent for users.

In one ERP project, this reduced manual approvals and simplified the overall workflow.

Yes, regularly. In product and B2B/ERP projects, I often translate business requirements into functional specs and UX documentation for development teams.

I structure user flows, edge cases, and interface states.

I define acceptance criteria together with PMs and engineers, and document component behavior and interaction logic.

I also collaborate with analysts and developers on API constraints and system limitations.

Tools include Figma, Jira, Confluence, User Stories, Job Stories, and flow diagrams.

In a B2B CRM project, I analyzed the effectiveness of an internal communication campaign promoting new features for managers and operators.

User segmentation was based on roles, behavioral data, feature adoption depth, and engagement levels.

Communication flows and UI hints were adapted for each segment.

Effectiveness was measured through open rate, CTR, adoption rate, usage frequency changes, CSI/CSAT, and support ticket reduction.

This approach improved engagement and made communication significantly more targeted.