about

I'm an AI/ML Engineer at Cloudly IO and a core contributor to Maveric, the Linux Foundation Connectivity platform for AI-native RAN optimization. My work spans the stack: I built the end-to-end MRO rApp pipeline (PPO-based, re-architected from linear to multi-threaded for a 45–55% runtime cut), hardened the Bayesian digital twin engine's caching and Kafka layers, wrote the 3GPP-compliant golden topology generators in the data simulator, and added observability across all five platform services plus the gateway's admin log plane.

Working day-to-day across the CloudlyNet AI platform, I also know its LangGraph multi-agent copilot architecture well.

I started my career in data annotation and QA, which taught me what production-grade training data actually looks like before I moved into ML engineering. When I'm not at the terminal, I'm reading research — the long-term goal is a role at an AI research lab.