Accelerating AI Workloads With Arrow Flight SQL on Teradata

What if the biggest bottleneck in your AI workflow isn't the model—it's getting data to it?

Many AI and analytics workflows across cloud, on-premises, and hybrid environments still rely on data movement, disconnected tools, and extra processing steps that add complexity and slow performance. As models grow larger and workloads become more distributed, efficient data access becomes critical.

Teradata's Arrow Flight SQL server endpoint provides high-performance, zero-copy access to data for AI and analytics workloads—delivering more than 24x performance improvement over JDBC and even higher gains in select scenarios. With native support for AI/ML, Python, PyArrow, Pandas, and model training workloads, it enables faster data access and model scoring while reducing the need for unnecessary data movement. Organizations can connect AI and analytics tools directly to Teradata data across cloud, on-premises, and hybrid environments.

Watch the live demonstration replay of how Arrow Flight SQL accelerates data access and model scoring workflows, helping teams move from business question to answer with less friction and stronger performance.

What you’ll see: 

  • A Python-based AI/ML workflow in Teradata AI Studio accessing data directly through Arrow Flight SQL, with no data movement and no extra steps 
  • High-performance model scoring against large datasets using native data access 
  • Real-world use cases, including distributed machine learning, data lake integration, LLM training, and real-time analytics 

Speakers include

Tamia van Geloven
Debjani Panda

Director of Product Management at Teradata

Debjani specializes in product development and strategy. She has extensive experience in leading product and program management across various industries and regions. Debjani holds an MBA from the University of Texas at Austin and a degree in electrical engineering from the National Institute of Technology in India.

계속 탐색