

AI agents in action
11 snips Jul 21, 2025
Swami Chandrasekaran, the Global Head of AI and Data Labs at KPMG, shares his insights on the powerful impact of AI agents in modern organizations. He discusses real-world success stories demonstrating how AI agents enhance executive decision-making and operational efficiency. Listeners will learn about the TACO Framework™, which categorizes AI agents by complexity and goal fulfillment. The conversation also covers governance, the importance of organizational readiness, and the balance between human and AI collaboration in driving innovation.
AI Snips
Chapters
Transcript
Episode notes
What Agents Actually Do
- Agents fulfill user-expressed goals by planning, reasoning, using tools, and taking actions autonomously.
- They combine language models, memory, tools, and orchestration to execute multi-step tasks without constant human prompting.
Classify Agents With TACO
- Use the TACO framework to classify agentic systems by goal and orchestration complexity.
- Design around Taskers, Automators, Collaborators, and Orchestrators to match goal complexity to architecture and governance.
Start With Taskers
- Organizations often start with Taskers because they produce quick pilots and measurable wins.
- Automators and Collaborators follow once teams grasp process reimagining and cross-functional orchestration needs.