Michał Kochaniak

Problem

Raw JTL and CSV outputs required hours of manual analysis, while reporting quality varied between cycles and stakeholders lacked self-service visibility.

Solution

Built an automated Python pipeline for performance analysis, chart generation, and structured DOCX/PDF reporting, with a local LLM layer for narrative summaries and anomaly flagging.

Problem

A hybrid banking app with native and webview flows required reliable automated regression across Android and iOS under frequent UI change.

Solution

Designed a layered automation framework with cross-platform abstractions, webview context handling, and CI-ready execution patterns.

Problem

Automated test results were disconnected from Jira workflows, making traceability, reporting, and quality visibility inconsistent.

Solution

Implemented a CI-integrated quality pipeline that publishes automated results to Zephyr Scale and maintains requirement-level traceability in Jira.

Problem

Senior QA time was repeatedly spent on routine tasks — reviewing logs, comparing baselines, and preparing reports.

Solution

Designed a locally hosted LLM-based assistant to support result interpretation, regression analysis, and QA reporting workflows in privacy-sensitive environments.