{"version":1,"type":"rich","provider_name":"Libsyn","provider_url":"https:\/\/www.libsyn.com","height":90,"width":600,"title":"AI Is Here: How Businesses Can Actually Use It w\/ Greg Whalen of Prove AI AZ TRT S07 EP17 (299) 10-11-2026","description":" AI Is Here: How Businesses Can Actually Use It w\/ Greg Whalen of Prove AI AZ TRT S07 EP17 (299) 10-11-2026   What We Learned This Week AI adoption is moving faster than the playbook. Businesses are experimenting with AI, but many still haven't figured out exactly where it belongs in their organization or who should be responsible for it. The next AI revolution may be about agents, not chatbots. The important shift could be from asking AI questions to having AI agents actually perform tasks and manage parts of a business workflow. Businesses will need to learn how to manage AI. The future may require employees and managers to work alongside AI agents, understand their limitations and evaluate their output. AI creates a new layer of business risk. Data, liability, security, legal responsibility and quality control all become important when AI is making decisions or performing work. The competitive advantage may be implementation\u2014not simply access to AI. Eventually, many companies may have access to similar AI models. The difference could be how effectively each company integrates AI into its people, processes, data and decision-making.         Guest: Greg Whalen \u2014 Prove AI Website: https:\/\/proveai.com\/ Greg is CTO at Prove AI, bringing a wealth of experience scaling startup and enterprise SaaS businesses. Greg is focused on leading our product development team through launch and successful adoption of our AI Governance application, Prove AI.   Greg has launched and scaled multiple startups and new enterprise product lines. His teams have invented and operationalized efficient AI\/ML components as part of these solutions across marketing, customer support, health care, and industrial sectors. Through his work applying AI to regulated use cases, he\u2019s passionate about finding efficient, pragmatic solutions to emerging needs of AI Governance. Under his leadership, his high-availability products have consistently grown revenue 5x-10x while exceeding desired service levels.   Greg brings a strong technical background and global leadership in product management, software development, and M&A. In the past, he has served as CTO for Xendit, a southeast Asia payments unicorn, GM for Amazon WorkMail, AWS\u2019s public email messaging service, the Asia, Pacific and Japan director for Pivotal\u2019s Data Science business, and product\/engineering leader (VP\/CTO) for Experian CheeahMail and Derm101.com. Prior to his work in software, he worked in AI\/ML research as a PhD candidate in Columbia University\u2019s Natural Language Processing group, where he published research on the use of machine learning for automated briefings of cardiothoracic surgeries. He holds a MS in Computer Science from Columbia University and a Bachelor of Arts in Computer Science and Music from Wesleyan University in Middletown, CT.   20+ years' experience in technology leading engineering and data platforms at global scale as CTO of Xendit, Southeast Asia\u2019s payments unicorn, General Manager for Amazon WorkMail at AWS, APAC Director for Pivotal\u2019s Data Science business, and VP CTO and product leader at Experian, CheetahMail, and Derm101.      The Vision for Prove.ai and AI for Business   Prove AI provides the foundation for predictable and reliable production AI models and agents. With Prove AI, builders can productionalize AI and prove its ROI by setting and maintaining custom performance metrics with their preferred tooling. Built on the open source OpenTelemetry and Prometheus projects, Prove AI allows builders to observe and troubleshoot while owning their data assets, improving mean time to remediation.   SHOW NOTES   SEGMENT 1 \u2014 How Should Businesses Handle AI? The Big Question AI is moving faster than many businesses can figure out how to manage it. Companies are asking two basic questions: How can we actually use AI to become more productive? How do we use AI responsibly and manage the risks? Many companies are experimenting with AI without a clear strategy or plan. Businesses are trying to identify: Realistic AI use cases Where AI can improve productivity Which employees should be responsible for implementing it How AI should fit into existing management structures Generative AI may represent another major technology revolution, similar to the Internet, software and cloud computing. The next evolution is AI agents that can actually perform work inside a business, rather than simply answering questions. This raises a bigger management question: Who owns AI inside the company? Will businesses eventually need an AI officer or AI leader? Who is responsible for managing the "AI team"? Businesses need to determine where AI belongs in their existing workflows rather than simply adopting AI because everyone else is doing it. Non-engineers can now build increasingly sophisticated AI applications using tools such as Claude Code. Engineers will still be important for defining specifications, architecture and requirements. Employees may increasingly need to be cross-trained to work with AI. Business leaders will need to understand AI well enough to make decisions about it. Eventually, customers may need to understand how businesses are using AI as well. Important Concept Every AI application ultimately involves: Inputs \u2192 Decision Process \u2192 Outputs The quality of the output depends heavily on how the system is designed, what information it receives and what rules govern the decision process. Host Question "Are businesses currently moving too quickly with AI without first figuring out what problem they're actually trying to solve?" SEGMENT 2 \u2014 How We Got Here & Where AI Is Going Greg's Background Greg began working in AI research in the late 1990s. At the time, computing power was limited, so much of AI was still theoretical. Early work involved developing algorithms that eventually became practical business applications, including marketing targeting. The growth of big data and analytics around 2014 accelerated the practical use of AI. Amazon's AWS became a major example of how cloud computing could provide enormous computing resources. AI and predictive analytics are now being applied to areas such as healthcare. Prove AI Prove AI is focused on helping businesses understand how AI can actually be used within their organizations. The challenge isn't simply: "Can AI do this?" The better question is: "Where does AI actually fit within the business?" That includes: Management Workflow Data Legal considerations Risk management Employee responsibilities Customer relationships Where AI May Be Going Foundation models such as ChatGPT and Claude will continue to improve. Businesses may increasingly develop smaller, specialized language models. Some systems may function more like retrieval\/RAG systems using a company's own information. Businesses could eventually have their own bespoke AI system acting somewhat like a junior analyst. Multiple AI models could work together to complete a larger task. AI tools may eventually function as a team of specialized digital workers. The Liability Question If several AI models are working together and something goes wrong: Who is responsible? Potential responsibility could involve: The business using the AI The developer of the software The AI model provider The person who designed the workflow The person who approved the output This creates a major legal and management issue that businesses are still working through. SEGMENT 3 \u2014 Crossing the AI Bridge What Prove AI Is Working On Prove AI is currently in the pre-product stage and is focused on helping companies cross the AI bridge. The challenge is partly technological\u2014but also human. The Human Side of AI Are employees open to change? Are managers prepared to manage AI? What happens when AI begins changing people's jobs? How do employees learn to work alongside AI agents? What skills will employees need? There isn't yet a universally accepted playbook for preparing a workforce for AI. The Future Workforce Instead of one employee doing every part of a task, businesses may eventually have: Employee + AI Agent + Specialized AI Agents A team of AI agents could potentially complete an entire business process. Employees and managers will need to learn how to: Give AI the right instructions Evaluate AI's work Understand AI limitations Manage AI agents Determine when human judgment is required Build workflows around AI The Bigger Issue People are understandably concerned because this is new territory. The question isn't simply: "Will AI replace people?" It may increasingly become: "How will people manage and work with AI?" SEGMENT 4 \u2014 Guardrails, Risk & Liability The Business Challenge Companies need better ways to establish: Rules Data sets Evaluations Controls Review processes Guardrails But businesses still don't always know what assurances they can expect from AI systems. Start With the Workflow One approach is to analyze the workflow first: What task is AI performing? What data is it using? What decisions is it making? What could go wrong? How do you evaluate the output? What rules should govern the system? Who is responsible when something goes wrong? Assume AI Can Fail A useful approach is to assume the AI system isn't automatically safe or reliable. This is similar to how cloud infrastructure is managed: Assume an external dependency can fail. The more powerful AI becomes, the more important it may be to have controls around it. The Legal Department's Concern One of the biggest questions will be: "Where are the guardrails?" Businesses need to know: What AI is allowed to do What data it can access What decisions it can make What it cannot do How its work is evaluated What happens when it makes a mistake This may require businesses to over-invest in controls early rather than discover problems after deployment. Human-in-the-Loop? Greg doesn't necessarily see the future as having a human manually review every AI decision. Instead, specialized AI systems may increasingly evaluate other AI systems using trusted or "golden" data. That could eventually create: AI doing the work \u2192 AI evaluating the work \u2192 Human managing the system BIG PICTURE \u2014 Where Is This All Going? AI is still developing rapidly. There are already powerful models from companies around the world, including: OpenAI Anthropic Chinese AI developers such as DeepSeek Other specialized and open-source models The technology will continue to evolve. But AI is still software. That means businesses can potentially manage it through: Good architecture Rules Data controls Testing Evaluations Risk analysis Governance Greg's Bigger View AI may eventually become a commodity technology, much like previous major technological advances. The competitive advantage may not simply come from having AI. It may come from knowing how to integrate AI into the business better than competitors.     Tech Topic: https:\/\/brt-show.libsyn.com\/category\/Tech-Startup-VC-Cybersecurity-Energy-Science  Best of Tech: https:\/\/brt-show.libsyn.com\/size\/5\/?search=best+of+tech   \u2018Best Of\u2019 Topic: https:\/\/brt-show.libsyn.com\/category\/Best+of+BRT      Thanks for Listening. Please Subscribe to the AZ TRT Podcast.     AZ Tech Roundtable 2.0 with Matt Battaglia The show where Entrepreneurs, Top Executives, Founders, and Investors come to share insights about the future of business.  AZ TRT 2.0 looks at the new trends in business, & how classic industries are evolving.  Common Topics Discussed: Startups, Founders, Funds & Venture Capital, Business, Entrepreneurship, Biotech, Blockchain \/ Crypto, Executive Comp, Investing, Stocks, Real Estate + Alternative Investments, and more\u2026    AZ TRT Podcast Home Page: http:\/\/aztrtshow.com\/ \u2018Best Of\u2019 AZ TRT Podcast: Click Here Podcast on Google: Click Here Podcast on Spotify: Click Here                    More Info: https:\/\/www.economicknight.com\/azpodcast\/ KFNX Info: https:\/\/1100kfnx.com\/weekend-featured-shows\/   ","author_name":"AZ Tech Roundtable 2.0","author_url":"https:\/\/brt-show.libsyn.com\/","html":"