Viktor Vasylkovskyi
Viktor Vasylkovskyi

Viktor Vasylkovskyi

Senior Software Engineer | AI Agents & LLM Systems

About Me

Senior Software Engineer with 9 years of professional experience including 2+ years of hands-on work in AI systems, specializing in LLM-based AI agents. Currently part of the AI team at PagerDuty, where I design and evaluate production-grade AI agents using state-of-the-art LLMs, lead MCP and agent-to-agent (A2A) integrations, and develop LLM evaluation pipelines (LLM-as-Judge for classification and reasoning tasks), as well as some of our core AI Agents. Currently seeking to deepen my research contributions in large language models, LLMs learning and training through PhD study.

My academic background includes a Master's in Computer Science and Engineering in IST, and have published peer-reviewed research on privacy-preserving architectures for human-robot interaction using blockchain, demonstrating experience in research, system design, and experimental validation.

Academic Background
Instituto Superior Técnico (IST) - Lisbon, Portugal

Master of Science in Computer Science and Engineering

Focus areas: Blockchain, Distributed Systems, Software and Network Security, Privacy in Human-Robot Interaction

Thesis Topic

Improving privacy in human-robot interaction through blockchain-based architectures, developed with Privacy by Design principles in a healthcare environment (Portuguese Institute of Oncology).

Thesis Overview

The recent interactions between humans and robots appear to have problems regarding humans privacy. The project's aim is to improve privacy in interaction between robots and humans using Blockchain technology. In other words, project is developed with Privacy by Design in mind. Health Care environment - IPO - Portuguese Institute of Oncology Technologies.

Open Source Contributions
Infrastructure as Code Toolbox - Practical AWS Infrastructure Education

Author and maintainer of Infrastructure as Code Toolbox, an open-source educational project demonstrating how to deploy modern applications securely on AWS. The project showcases production-grade infrastructure patterns using Terraform and infrastructure-as-code best practices.

This project provides deep hands-on experience in cloud architecture, DevOps practices, and infrastructure automation - skills directly applicable to building and scaling production systems.