My unique edge lies in bridging the gap between deep model research and scalable AI production. Over the past two years, I’ve immersed myself in ML and Generative AI, culminating in my recent role as the founding Full Stack AI Engineer at krid.ai. Working directly alongside the technical co-founder, I architected and deployed complex, multi-agent orchestration systems from scratch. This included building voice and WhatsApp agents, custom round-robin RAG pipelines, and a 'Talk to My Data' agent capable of querying hundreds of tables and Agent Evaluations over GCP. I also led a team of 4 AI interns and helped scale the company’s infrastructure from 1 to 10 clients, personally managing two of them.
Parallel to this, my master’s research focuses on the extreme opposite of massive cloud deployments: Edge AI. I successfully adapted a Tiny Recursive Model (TRM) for text generation using continuous diffusion techniques. By keeping the model at just 19 million parameters with a 72 MB memory footprint, I established a highly promising direction for running generative tasks on heavily constrained devices.
Whether it’s optimizing small language models for edge hardware or orchestrating cloud-based multi-agent systems, I bring a rare blend of rigorous AI research and rapid, startup-paced engineering to the table.