I am a Construction-Manager-turned-Senior-Test-Automation-Engineer with a deep focus on production efficiency and system throughput. My mission is to bring the rigor of enterprise test automation to non-deterministic AI systems: testing LLM-powered agents, guarding their outputs, and building quality strategy for cross-functional teams shipping AI features.
- Modern Automation: Architecting robust, isolated, and parallelized testing frameworks using .NET (C#) and Playwright.
- Clean Architecture: I strongly advocate for Playwright best practices—utilizing Locator APIs over raw selectors, relying on native auto-waiting instead of
Thread.Sleep, and designing small, reusable components over monolithic Page Objects. - AI Agent Development (Python): Building tested AI agents with Azure AI Foundry, Azure OpenAI, Semantic Kernel (plugins, memory), Bot Framework, and Microsoft Graph enterprise data connectors. Every project ships with a real pytest suite.
- Testing & Securing AI Code: Unit and integration testing of LangChain-based code with pytest, mocking LLM responses, coverage targets, and dependency/vulnerability assessment for AI frameworks.
- LLM Output Safeguarding: Turning safety policies into automated behavioral tests for LLM outputs, plus mutation testing (
mutmut) to prove the tests actually catch regressions. - LangChain Agents: Tool calling and chaining, LCEL, and LangChain's built-in agents.
- DevSecOps & Infrastructure: CI/CD with Jenkins and GitHub Actions. Exploring modern security paradigms, Zero-Trust policies, supply chain mitigation, and secure secret management via Azure and HashiCorp Vaults.
- 🔄 Agentic AI with LangChain & LangGraph (IBM): stateful graph workflows, memory, Reflection/Reflexion/ReAct agents, multi-agent systems, and agentic RAG.
- 🧪 Next up: an LLM evaluation harness, grading real model output for factuality, format adherence, and regressions instead of mocking it away.
- Microsoft: AI Agent Fundamentals with Azure AI Foundry
- Microsoft: Code and Framework-Based Agent Development
- Coursera: Test and Secure Your AI Code
- Coursera: Safeguard LLM Outputs: Test and Evaluate
- IBM: Fundamentals of Building AI Agents
C# .NET Playwright Selenium NUnit Python pytest mutmut Azure AI Foundry Azure OpenAI Semantic Kernel Bot Framework Microsoft Graph LangChain LangGraph (learning) Jenkins GitHub Actions
- ⛰️ Pushing limits through mountain ultra-marathons and CrossFit.
- 🎸 Coding to a solid metal soundtrack (From Ashes to New, Amaranthe, Beyond the Black, Evanescence).
- 📚 Recharging with sci-fi, horror/adventure movies, and the dark fantasy worlds of Vasil Popov.
- Email: v.vakareev@gmail.com
- Collaboration: Always open to discussing AI quality and agent testing, LLM evaluation, scalable Playwright architectures, or the best trail running gear!