I’m currently working on
Designing enterprise-grade agentic AI systems — multi-agent orchestration, evaluation-first workflows, and RL-based environments that make tool-using LLMs reliable in production.
I’m looking to collaborate on
Agentic applications, structured intent → tool mapping systems, reinforcement learning for LLM reliability, and production-scale GenAI platforms.
I’m looking for help with
Advancing large-scale evaluation frameworks for multi-step agents and pushing the boundaries of RL + synthetic environments for structured reasoning.
I’m currently learning
Advanced LLMOps, distributed inference optimization, and deeper reward modeling strategies for agent generalization.
Ask me about
Multi-agent architectures, RAG vs memory trade-offs, failure modes in tool-using agents, personalization at 100M+ scale, or how to take GenAI from POC to production safely.
Fun fact
I’ve built AI systems that process 10K+ alerts daily — and I still get excited debugging why an agent decided to call the wrong tool.
