TestCon Europe 2026
October 20-23
Vilnius & Online
Oleksii Bakunin
About
With “13 years” of experience in software quality and delivery, “Oleksii Bakunin specializes” in building teams, processes, and automation systems that help companies ship products faster and with confidence. His journey has taken him across “FinTech” (payments “and” lending), “EdTech,” Travel, and “VoIP,” where he led teams and initiatives that transformed engineering organizations.
Key strengths:
Leadership: Built QA teams from scratch, scaled departments, and mentored engineers to become high-performing contributors.
Process architecture: Established QA and release processes that integrated seamlessly with product and engineering, reducing risk and increasing velocity.
Automation at scale: Designed end-to-end automation frameworks “(Web, Mobile, API)” and integrated them with “CI/CD” pipelines, making release cycles predictable and resilient.
“AI integration:” Introduced “AI” into daily QA routines to build automated pipelines.
Cross-domain expertise: Hands-on experience with cloud-based, client-server, and mobile platforms, combined with a deep understanding of “FinTech” reliability requirements.
Oleksii Bakunin | Al-Augmented QA in Practice: How To Build LLM-Powered QA Workflow without loosing control
In this talk, Oleksii Bakunin will show how to make test documentation AI-friendly and how teams can implement a self-updating test documentation workflow while maintaining control.
After attending this session, participants will be able to:
Design AI-friendly test guides that chunk well and stay readable for humans.
Build a closed-loop workflow: a ticket with requirements → AI-generated test instructions and prompts for developers to use while writing a feature → QA approval → documentation update.
Avoid common RAG failures “such as” bad chunking, missing context, duplicated knowledge, “and” outdated docs.
Introduce AI into QA without losing control, using approval, confidence signals, and versioning.
Apply the approach to any product, even where documentation is currently messy or incomplete.