Hello! I'm Sina, an AI engineer and researcher driven by a fascination with how machines can learn to perceive and interact with the world in complex, human-like ways. My work is focused on turning cutting-edge research into practical, real-world applications that solve meaningful problems.

My journey began in academia, where my Ph.D. research explored the intersection of vision and language. I developed novel deep learning models that could answer complex questions by analyzing text and images. I have also worked on controlling robotic systems using natural language and have built powerful multimodal prediction systems that have been tested on diverse geo-spatiotemporal datasets for tasks like weather forecasting and traffic prediction.

I've since carried that passion into the industry. At American Airlines, I led the creation of a forecasting engine that cut flight traffic prediction errors in half. Later, at Truveta, I had the opportunity to contribute to the mission of "Saving Lives with Data" by building NLP models that processed sensitive clinical data, significantly accelerating research workflows.

Currently, as a Lead AI Engineer at Armada, I manage a team of AI engineers building the next generation of intelligent industrial systems. I lead the development of a sophisticated agentic AI platform that delivers real-time, natural-language insights to operators on the ground. Our system, optimized for edge deployment, already outperforms larger proprietary models by over 10% and is on track to boost industrial efficiency by 4.5%.

Alongside this, I lead our AI safety and red-teaming research. Our latest work studies how open-weight LLMs can be turned into "sleeper agents" — models fine-tuned to hide malicious tool-calling behavior behind benign benchmark performance — and develops layered defenses that reliably detect them. This research is currently under review at NeurIPS 2026. My broader research record includes 10 publications and 5 U.S. patents spanning multimodal learning, LLM post-training, and agentic AI.

I believe in building AI that is not only powerful but also collaborative and accessible. I'm a strong advocate for open-source and am always excited by the challenge of creating robust, scalable solutions that empower people and drive innovation.