Senior Software Engineer @ Microsoft
GenAI Engineer building production LLM systems, enterprise AI infrastructure, and agentic platforms.
I work at the intersection of LLM applications, agent protocols, enterprise RAG, and AI for operations — turning prototypes into systems that are testable, maintainable, and safe enough for real users.
- Production LLM systems — RAG, tool calling, guardrails, evaluation, observability, and lifecycle design.
- Agent infrastructure — MCP, multi-step workflows, memory, typed tools, and failure-aware orchestration.
- Enterprise Java AI — bringing modern AI capabilities into Spring Boot / JVM ecosystems.
- AI for DevOps & SRE — Kubernetes diagnosis, incident reasoning, Git/Jira correlation, and safe remediation loops.
I contribute to the LangChain4j ecosystem with a focus on infrastructure-level work that expands what Java AI applications can do in production.
Led and delivered key pieces of the Java MCP server path across LangChain4j and LangChain4j Community:
- Shared MCP protocol DTOs and stdio JSON-RPC transport foundations.
- Community MCP server module exposing Java
@Toolmethods over MCP. - Java stdio MCP server example and documentation for Claude Desktop-style usage.
- Lifecycle hardening such as clean stdio shutdown support.
Built agent-ready integrations that connect LLMs to enterprise systems:
- Jira Tool — search, create, comment, ADF parsing, robust agent-facing errors.
- Web Scraper Tool — lightweight HTML-to-Markdown extraction with context-noise reduction.
- Confluence / GitLab Document Loaders — resilient enterprise knowledge ingestion for RAG.
- Built prompt repetition components for AI Services and RAG: policies, input guardrails, query transformers, AUTO-mode gates, idempotence, and docs.
- Fixed LangChain4j AI Services guardrail ordering so multimodal input is materialized before guardrail execution.