Андрій

Andrii Vozniuk
QA Automation Engineer · Security & AI Tooling
Tel: [відкрити контакти](див. вище в блоці «контактна інформація») · [відкрити контакти](див. вище в блоці «контактна інформація») · github.com/aovozniuk1 · Kyiv, Ukraine

Qualifications
Objective: QA Automation Engineer focused on test-automation architecture and security-test automation.
Languages: Ukrainian (native), English (Upper-Intermediate)
Programming: Java, Python, SQL, Bash
Core QA Automation: Selenium WebDriver (Java), Playwright, TestNG, JUnit, Rest Assured, Page Object pattern, Appium,
Selenide, Selenoid, Gatling, WireMock, Postman, REST/API testing, Contract testing (Pact), Saucelabs, Allure
CI/CD & Tooling: Maven, Git, Jenkins, TeamCity, Docker, Docker Compose, GitLab CI, Jira, TFS
Cybersecurity / Pentesting: DAST and authenticated web/API security testing, vulnerability scanning, exploit-chain synthesis
with live replay; IDOR/BOLA, SQLi/XSS/SSRF/XXE, GraphQL security; OWASP Top 10; MCP server security; WAF-resilient
scanning
AI / LLM Engineering: Multi-provider LLM abstraction (cloud + local models), prompt engineering & caching, hybrid routing with
cost tracking; ReAct / tool-calling agents, structured output, RAG, agent memory & orchestration; LLM/AI evaluation (eval design,
LLM-as-a-judge, RAG groundedness)

Experience
N-iX — QA Automation Engineer
2024 – August 2026
• Build and maintain Java AQA frameworks (Selenium WebDriver, TestNG, Rest Assured) across UI and API layers using the
Page Object pattern; integrate suites into CI/CD with Allure reporting
• Drive cybersecurity-focused QA — API and web security testing aligned with OWASP Top 10, plus automated security scanning
in the pipeline
Own project — Pentkit
2026 – Present
• Author of Pentkit, an AI-native DAST / pentest toolkit — 140+ scanners, 55+ recon modules, 80+ payload generators, with a
multi-agent AI layer for authenticated recon, scanning, exploit-chain proof and reporting
• Designed Pentkit's AI infrastructure — multi-provider LLM layer with caching and hybrid routing, plus a multi-agent analysis
pipeline (ReAct investigation, skeptical auditor, exploit-chain synthesis with live-replay validation)
Own project — LLM eval harness
2026
• Built a pytest-based evaluation harness for LLM features: reference-based (exact / semantic), reference-free (structural,
guardrails) and LLM-as-a-judge scorers; accuracy, precision/recall/F1 and pass-rate gates over repeated runs, wired as a CI
quality gate
• RAG evaluation — retrieval hit / context precision, groundedness (anti-hallucination) and abstain-when-not-in-context checks;
same suite runs against a stub or a real model. Companion lab: consumer-driven contract tests (Pact + Pact Broker in Docker
Compose)
Euristiq — QA Automation Engineer
February 2022 – 2024
• Developed from scratch and maintained the API and UI test frameworks
• Configured and maintained integration of AQA tests with Saucelabs
• Mentored the manual QA team
Form.com — QA Automation Engineer
December 2018 – February 2022
• Updated the existing autotest framework; set up and configured Selenoid
• Configured TeamCity build chains
• Wrote a test framework from scratch using WireMock
• Manual, web, and load testing across projects
Genesis — QA Automation Engineer
January 2018 – December 2018
• Updated the framework for automated testing of rabota.nur.kz
• Created an Appium framework from scratch for the mobile app
• Set up Jenkins from scratch and configured jobs
• Manual layout testing
CyberianSoft — QA Automation Engineer
January 2017 – January 2018
• Participated in building a framework for automated testing of the AMT back office (Java, Selenium WebDriver, TestNG, Maven)

Courses
Prog.kiev.ua — QA Automation, QA Manual (2016) · IT Labs — Advanced Java (2014)