* Architected and deployed self-hosted LLM infrastructure for production-oriented AI workloads, including model serving, API integration, GPU resource management, and inference optimization.
* Designed multi-agent architectures for agentic software engineering workflows, orchestrating specialized AI agents for code generation, debugging, testing, documentation, and automation.
* Integrated LLM-based agents and developer tools into CI/CD pipelines to automate software production, validation, testing, and deployment workflows.
* Developed automation workflows combining LLMs, agentic tools, development environments, and CI/CD infrastructure to improve engineering productivity and reduce manual effort.
* Evaluated LLM models, inference frameworks, and agent orchestration approaches for reliability, tool use, scalability, and production deployment.