Amit Marathe — Director of Engineering

I am passionate about leading geographically distributed ML teams to build and scale innovative multimodal LLM and GenAI solutions. My focus is on enabling the infrastructure that allows these technologies to solve real-world problems and make a difference. I am dedicated to bringing cutting-edge LLM research to life through practical applications. My focus is on bridging the gap between research and production for LLMs, leading the development of end-to-end infrastructure for productionization, from pre-training to launch governance, including fine-tuning tools, evaluation tools and multi-agent systems. I have experience in optimizing LLM serving and training efficiency using various techniques, including compiler and scheduler optimizations, quantization, and other inference optimization methods I've collaborated closely with Google DeepMind on several initiatives, bringing research advancements into production and contributing to the development of cutting-edge GenAI models. I'm a strong leader with experience in building and leading high-performing global teams of ML engineers, SWEs and TPMs, driving successful project delivery and business outcomes. I believe in fostering a collaborative engineering culture, which is why I built the BrickML community, connecting over 1,500 Googlers with hands-on experience across the end-to-end ML stack. I also lead hackathons to drive innovation and boost team morale, encouraging everyone to learn, experiment, and grow. Key Skills & Technologies: GenAI & LLMs: LLM productionization, model training and deployment, pre-training infrastructure, fine-tuning tools, evaluation tools, agents, multi agent systems, launch governance Cloud Platforms: Google Cloud, AWS, Azure Software: Python, C++, TensorFlow, PyTorch, Jax, scikit-learn, Kubernetes, Docker, Data Engineering: Spark, Hadoop, Synapse, AzureML

Stackforce AI infers this person is a leader in AI and ML infrastructure development.

Location: San Francisco, California, United States

Experience: 25 yrs