AI engineer building production multi agent systems. MCP servers, RAG pipelines, voice agents, and open source eval tooling that survives real users.
I am an AI engineer building production multi agent systems. My focus is the layer between architecture and deployment where most AI projects stall: MCP servers, RAG pipelines on knowledge graphs, Pydantic validated agent workflows, and observability that catches regressions before users do.
I have learned from working across data analytics, program analysis, and production AI that the hard part is not building models. It is building solutions that people actually adopt, trust, and rely on in real environments where things break.
Currently building multi-agent architectures where GenAI moves from demos to production. I am deep in this space because we are still in the early innings of figuring out what these tools can really do, not the hype stuff, but the practical applications that matter.
Building production-scale multi-agent AI systems and GenAI applications at an AI-first startup. Architected the full backend from scratch.
Led data analytics initiatives for a global e-commerce company, driving business growth through data-driven insights.
Built data infrastructure and ML pipelines for enterprise clients, establishing the foundation for advanced analytics.
GenAI tool using LLMs to generate personalised, context-aware cold emails from a job page link. Uses prompt engineering and few-shot learning for high-quality outreach at scale.
View on GitHub →Processed time-series fMRI data to classify subjects using deep learning. Compared LSTM, CNN, and SVM approaches. LSTM achieved 76% accuracy on the benchmark dataset.
View on GitHub →ETL processes and real-time anomaly detection for financial market data. Designed data quality checks and dashboards for monitoring and reporting across multiple data sources.
View on GitHub →I love talking about multi-agent systems, production AI, and how to take cool demos and make them ship. Reach out.