I invited Kyle Hauptfleisch, Chief Growth Officer at Daemon, to strip the buzzwords out of AI and talk plainly about what moves the needle at work. The conversation began with an honest look at why so many pilots stall. It ended with a calm, workable path for leaders who want results they can measure rather than demos that gather dust. Along the way we compared two very different mindsets for adoption, AI added and AI first, and what that means for teams, accountability, and the way work actually gets done.

Here’s the thing. Plenty of organisations raced into proofs of concept because a board memo said they had to. Kyle has seen that pattern play out for years, and he argues for a simpler starting point. You do not need an AI strategy in a vacuum. You need a business strategy that names real constraints and outcomes, then you pick the right kind of AI to serve that plan.

AI Added vs AI First

This distinction matters. AI added means dropping tools into the current way of working. Think code generation that saves hours on day one, only to lose those hours later in testing, release, or approvals. The local gain never flows through to the customer.

AI first asks a harder question. How do we change the workflow so those gains survive from whiteboard to production? That can mean new handoffs, fresh definitions of ownership, and different review gates. It is less about tools, more about the shape of the system they live in.

Accountability sits at the center. Kyle raised a scenario where a lead might one day direct fifty software agents. The intent behind those agents remains human. So does the responsibility. Until structures reflect that, companies will cap the value they can safely realise.

From Pilots to Production

Kyle offered a simple mental model that avoids endless experimentation. Picture a Venn diagram with three circles. First, a real constraint that people feel every week. Second, usefulness, meaning AI can change the outcome in a measurable way. Third, compartmentalisation, so the work sits far enough from core risk to move fast through governance. Where those circles overlap, you have a candidate to run live.

He shared a small but telling example from Daemon. Engineers dislike writing case studies after long projects. The team now records a short conversation, transcribes it with Gemini inside a safe, private setup, and drafts the case study from that transcript. People still edit, but the heavy lift is gone. It saves time, produces more human stories, and proves a pattern the business can repeat.

Leaders can start there. Pick a contained problem, run it in production, measure the outcome, and tell the truth about the bumps. That story buys trust for the next step, which is how you scale without inflating the promise.

Humans, Accountability, and Culture

We talked about the fear that AI erases the human role. Kyle’s view is steady. Models process data. People set intent, judge context, and carry the can when decisions matter. Agents will take on more tasks. The duty to decide will remain with us.

Upskilling then becomes less about turning everyone into a prompt whisperer forever and more about teaching teams to think with these tools. Inputs improve, outputs improve. Middle managers, in particular, gain new leverage for research, planning, and option testing. The job shifts toward framing better questions and challenging the first answer that comes back.

In this episode I sit down with Mo Cherif, Vice President of AI Innovation at Sitecore, to explore one of the biggest shifts in business today: the rise of agentic AI. Unlike traditional AI models that focus on narrow tasks, agentic AI brings autonomy, reasoning, and collaboration between specialized agents. It is changing the conversation from automation to transformation.

Mo explains how agentic AI is reshaping marketing, customer engagement, and creativity. From hyper-personalized chat-driven discovery to removing repetitive project management tasks, we look at how AI can free marketers to focus on strategy, storytelling, and innovation. He also shares why success depends on three foundations: context, mindset, and governance.

We dig into Sitecore’s three pillars of brand-aware AI, co-pilots, and agentic orchestration, and how the company’s AI Innovation Lab, launched with Microsoft, helps brands experiment, co-innovate, and apply these ideas in practice. Mo also reflects on lessons from real projects such as Nestlé’s brand assistant and looks ahead to a future where personal AI agents interact directly with others on our behalf.

If you want to understand how agentic AI is moving from hype to real business impact, this episode will give you practical insight into what is already happening and what comes next.

*********

Visit the Sponsor of Tech Talks Network:

Land your first job in tech in 6 months as a Software QA Engineering Bootcamp with Careerist

https://crst.co/OGCLA

When access to advanced AI models is no longer the big differentiator, the real advantage comes from how effectively a business can connect those models to its own unique data. That was the central theme of my conversation with Rahul Pathak, Vice President of Data and AI Go-to-Market at AWS, recorded live at the AWS Summit in London.

In a bustling booth on the show floor, Rahul explained how AWS is helping organisations move from AI pilots to production at scale. We discussed the layers of infrastructure AWS provides, from custom silicon like Trainium and Inferentia to services such as SageMaker, Bedrock, and Q Developer, and how these combine to give enterprises the flexibility and performance they need to build impactful AI applications.

Rahul shared examples from BT Group, SAP, and Lonely Planet, each showing how the right blend of tools, data, and strategy can lead to measurable business results. Whether it is accelerating code generation, generating custom travel guides in seconds, or using generative AI to produce personalised content, the common thread is a focus on business outcomes rather than technology for its own sake.

A key point in our discussion was that most companies do not have their data ready to power AI effectively. Rahul broke down how AWS is helping unify siloed data and make it available to intelligent applications, turning a company’s proprietary knowledge into a competitive edge. We also touched on responsible AI, sustainability, and the operational challenges that come with scaling AI, from cost efficiency to security and trust.

For leaders still weighing up whether to invest in generative AI, Rahul’s message was clear: waiting too long could mean being left behind. This episode is a practical guide to what it takes to deploy AI with purpose and how to ensure it delivers lasting value in a fast-changing market.