This episode of the Eye on AI podcast is sponsored by JLL.
JLL's AI solutions are transforming the real estate landscape, accelerating growth, streamlining operations and unlocking hidden value in properties and portfolios. From predictive analytics to intelligent automation, JLL is creating smarter buildings, more efficient workplaces and sustainable cities.
To learn more about JLL and AI, visit: jll.com/AI
In this episode of the *Eye on AI* podcast, we explore the world of AI at Google Cloud with Nenshad Bardoliwalla, Director of Product Management for Vertex AI.
Nenshad unpacks the three core layers of Vertex AI: the Model Garden, where users can access and evaluate a diverse range of models; the Model Builder, which supports model fine-tuning and prompt optimization; and the Agent Builder, designed to develop AI agents that can perform complex, goal-oriented tasks.
He shares insights into model evaluation strategies, the role of Google’s Tensor Processing Units (TPUs) in scaling AI infrastructure, and how enterprises can choose the right models based on performance, cost, and regulatory requirements.
Nenshad also delves into the challenges and opportunities of AI prompt optimization, highlighting Google’s approach to ensuring consistent outputs across different models. He discusses the ethical considerations in AI design, emphasizing the need for human oversight and clear guardrails to maintain safety.
Whether you’re in AI, tech, or curious about AI's potential impact, this episode is packed with insights on next-gen AI deployment.
Don’t forget to like, subscribe, and turn on notifications for more episodes!
Stay Updated:
Craig Smith Twitter: https://twitter.com/craigss
Eye on A.I. Twitter: https://twitter.com/EyeOn_AI
(00:00) Introduction to Nenshad Bardoliwalla & Vertex AI
(01:52) Overview of Vertex AI's Three Core Layers
(05:35) Nenshad's Journey to Google Cloud
(06:36) Choosing the Right AI Model
(08:00) Google’s AI Infrastructure & Tensor Processing Units (TPUs)
(10:15) Model Builder: Fine-Tuning & Prompt Optimization
(12:11) Agent Builder: Building AI Agents with Tools & Planning
(17:57) Model Evaluation & Prompt Management
(21:23) Generative AI for Business Analysts
(23:24) AI Model Modality & Use Case Selection
(25:23) Popularity Distribution of AI Models
(28:18) Prompt Optimization Tools
(34:20) Building AI Agents: Real-World Use Cases & Ethical Safeguards
(40:13) The Capabilities & Limitations of AI Agents
(45:48) TPU vs. GPU
(50:33) Future of AI at Google Cloud