Tech Talks Daily

When would you trust your ERP system to move from answering a question to carrying out the work?
Recording at SuiteWorld in Las Vegas, I spoke with Brian Chess, who leads AI product and technology for Oracle NetSuite, about what that progression looks like in practice. Brian was an early developer when the business was called NetLedger, later left to build a company, and returned with a different perspective on the needs of small and growing businesses. Our conversation focuses on how AI can help people understand their systems, complete unfamiliar tasks, and gradually delegate work without losing control of the process.
One of Brian's most useful ideas is AI acting as a teacher. He describes users asking Ask Oracle who has logged into their account, or asking it to explain a script they do not fully understand. The benefit is the answer and the explanation of how to find it. Someone who previously needed help from a colleague can begin to understand a part of the system for themselves.
He also shares an example involving NetSuite customer MCNC. When the payroll administrator was unavailable, Brian says a controller who had not been trained on the process used Ask Oracle for guidance. The outcome he describes is straightforward and meaningful: payroll ran on time. It is a customer account from the interview, rather than an independently assessed result, but it illustrates how useful AI assistance can be without requiring a dramatic productivity claim.
Financial close provides another practical example. Brian describes starting with the spreadsheet a finance team already uses, modelling that process inside NetSuite, and introducing automation in stages. I recognise the appeal of that approach from my own time supporting month-end and year-end work in IT. Teams need to understand how the process will run, especially when a mistake can affect reporting or the system of record.
We discuss how permissions apply as software gains the ability to act. Brian says Ask Oracle operates with the permissions of the user, while an agent receives an assigned role governing its access. He describes a progression from read-only analysis to limited changes as users gain confidence. Human approval remains available for transactions, particularly while teams refine instructions and account for cases they may have missed.
Business context and shared knowledge are part of the same discussion. Brian explains how written skills can capture procedures and the intended business outcome, making knowledge available beyond the colleague who usually handles the task. We also talk about accountability. He describes transaction records showing which agent acted and when, together with an activity log. Those records help establish what happened, although they should not be confused with a complete explanation of every internal model decision.
The conversation also covers whether to build AI capabilities or use those embedded in a business platform. Brian's answer depends on the role the capability plays in a company's competitive position, the expertise available, and the time it wants to invest. His view of ROI is similarly specific to the business. Comparing vendor contracts with billing records may reveal savings, but the same result does not necessarily transfer to a company with different suppliers and arrangements.
We finish with the opportunity Brian sees for smaller businesses that have fewer specialists available. AI assistance may help people undertake work they previously needed a consultant or experienced colleague to explain. The question is how to extend that access while retaining appropriate permissions, review, and responsibility.
I would love to hear your thoughts, so please share your experiences. Where would conversational ERP make the biggest difference in your business, and which actions would you still want a person to approve?

What if the biggest gain from AI comes from reducing the distance between the person who understands a problem and the person building its solution?
On Tech Talks Daily, I'm joined by Ben Grinnell, Chief AI and Capabilities Officer at Newton, a strategic delivery consultancy working with private and public sector organizations. Ben's career stretches from an early web consultancy to enterprise technology delivery and today's AI tools. He describes returning from retirement after reading a draft book about vibe coding and realizing he wanted to be part of what came next.
What interested me was his practical explanation of how AI can change delivery. When a consultant or business specialist can test an idea while hearing the problem directly, fewer handoffs may mean less lost context and quicker feedback. Ben sees an opportunity to build useful solutions closer to the work. He also warns against confusing a demonstration that looks finished with software that is ready to run reliably. Security, architecture, testing, and maintainability still need people who know which questions to ask.
We discuss the relationship between business teams and IT. Employees have been building spreadsheets, macros, and local workarounds for years, so the challenge of technology appearing outside central oversight is familiar. AI changes the speed and scope of that activity. Ben argues for keeping technical expertise close to the business, connecting small projects to existing enterprise investments, and producing value early enough that people want to participate.
Trust runs through the conversation. Ben describes a progression from private use of AI assistants to shared agents that appear within a team's existing tools. Visibility can help colleagues understand what is happening, but easier access to company information can also expose old permissions problems. A document that has sat unnoticed on a shared site can become easier to find, making sensible access controls part of adoption rather than an afterthought.
One of the more surprising examples concerns legacy modernization. Drawing on his experience with caseworkers using different workarounds, Ben discusses separating a technical platform replacement from changes to how people work. He describes AI-assisted rewriting that preserves existing behavior before a later, more gradual process redesign. The conversation presents this as an approach to consider, rather than a demonstrated deployment or a reason to preserve security defects.
We also examine AI measurement. Ben challenges the emphasis on how many roles a company can remove and suggests looking at revenue, client service, profitability, and the repetitive work people no longer have to do. His argument is that the way a program is described affects whether employees support it. Leaders still need evidence that a project works, alongside an understanding of how their success measures influence behavior.
We finish with Newton's hands-on AI work, including internal hackathons, model access, and Ben's account of investment in sovereign compute for inference costs and UK data requirements. He makes a useful distinction between giving new users a strong first experience and asking experienced users to improve efficiency. Cost control and adoption need to be considered together.
I'd welcome your thoughts. How is your organization measuring AI value, and which small improvement has persuaded people to change the way they work?

What if your ERP system came to you through a conversation, instead of asking you to learn its menus before you could get anything done?
Recording at SuiteWorld in Las Vegas, I welcomed Evan Goldberg, founder and executive vice president of NetSuite at Oracle, back to the podcast. We last spoke two years earlier about the company's origins. This time, the discussion centres on his description of AI as the UI and what that could mean for the people using business software every day.
Evan describes an ERP experience in which users explain what they want in natural language. One example is adding a dashboard metric by describing the information it should show. If the result is not quite right, the user can explain what needs to change. For me, the useful question is whether that makes software easier to fit around a person's work and how reliably it understands the request.
We discuss the knowledge behind those answers. Evan describes SuiteSkills as a way to bring together information about NetSuite and the procedures businesses use within it. He connects that work with consistent answers through NetSuite Next and through external tools using the AI connector service. His ambition is to support the models and ways of working customers choose, while making the information returned from NetSuite dependable. It is a product direction he describes, rather than an independently tested guarantee that every model will behave identically.
The conversation then moves from answers to actions. Evan points to order approvals, accruals, and agents looking for exceptions that deserve attention. He argues for beginning with a process people already want to improve and deciding which parts can run with human approval or greater autonomy. The appropriate balance will depend on the company and the work involved. A conversational interface does not remove the need to decide where authority belongs.
Personalisation is another part of adoption. Evan's view is that people are more likely to keep using an AI feature when they can explain how they want it to work. The dashboard example makes that concrete, but it also raises a wider question about how companies maintain consistent procedures while allowing people to adapt their tools.
We also discuss smaller businesses and the opportunity to access capabilities without building a large specialist team. Evan describes a startup offering and his belief that the same AI models can help companies of different sizes. The practical considerations still include eligibility, the work a team needs to accomplish, and the controls around its business information.
For partners, he sees an opportunity to make industry expertise available through reusable skills. A specialist's knowledge of a particular niche could inform how Ask Oracle helps customers, while the partner spends time developing that expertise further. It is an interesting view of how advice and software configuration may change as AI becomes part of delivery.
I also ask about software pricing. If agents complete tasks that once required someone to log in, what happens to the traditional per-user model? Evan acknowledges the question but offers a counterpoint: easier access to information and personal agents could encourage people to remain users. He does not announce a pricing change, leaving us with a business-model question that is still developing.
We finish with an unexpected customer conversation at Disneyland, which Evan uses to illustrate interest from everyday business users. It is a memorable anecdote rather than a measure of adoption, but it brings the discussion back to the people who ultimately decide whether these features earn a place in their working day.
I would love to hear your thoughts, so please share your experiences. If AI becomes the interface to your business systems, what would you want it to understand before you let it act on your behalf?

What needs to happen before AI agents can reliably complete work across the systems that run a business?
In this episode of Tech Talks Daily, I welcome back Irfan Khan, Chief Product Officer for SAP Data & Analytics, to discuss the distance between a useful AI pilot and a dependable operational workflow. Our conversation begins with the autonomous enterprise and what that means for teams working across finance, procurement, supply chain and HR. Irfan explains why agents need a shared understanding of business activity, including the state of purchase orders and inventory, before coordination can produce a useful result.
We discuss the data quality, permissions and governance questions that become harder to manage as deployments grow. Irfan also considers how leaders should measure AI performance. Generating additional content or consuming additional tokens tells us something about activity, but leaders still need to understand the effect on productivity, accuracy and the cost of completing work. He shares how SAP has approached developer experimentation and why teams need to become deliberate about which tasks benefit from AI consumption.
The conversation also covers company memory, an initiative Irfan describes as a way to bring institutional process knowledge into agent workflows. We discuss how employee trust depends on the task, why order entry requires a different level of precision from a forecast, and how open architecture can preserve choices as models and infrastructure develop. Irfan then explains SAP's approach to connecting SAP and non-SAP data through SAP Business Data Cloud and the wider SAP Business AI Platform.
For leaders assessing the next stage of their AI programs, the discussion offers practical questions about context, access, cost and the business outcomes they want to improve. You can find the resources discussed in the accompanying show notes at TechTalksNetwork.com. Where is your organization finding the biggest obstacle to operational AI, and what would help you move forward?

When AI makes writing code faster, how do you decide whether your team is building the right thing?
In this episode of Tech Talks Daily, I speak with Johnny Halife, CTO at Southworks, about what happens to software engineering when code generation becomes easier but decisions still require experience. The pressure to deliver faster is familiar. The harder question is whether teams have the review capacity, testing processes, and deployment infrastructure to turn that extra output into useful software.
Johnny describes a different way of directing AI coding agents. Engineers need to specify the outcome, define what finished work should look like, and provide checks that can show whether it meets the required standard. Constantly telling an assistant what to do next can keep a person busy without giving the system enough responsibility to complete a coherent piece of work. We discuss how to create tasks that are small enough to review and substantial enough to delegate.
We also talk about context. Giving an AI system access to institutional knowledge sounds sensible, but adding every document to every conversation creates its own problems. Johnny describes borrowing from open source contribution practices, including README files, contribution guidelines, and records of architectural decisions. His approach combines an explanation of why the system works as it does with directions for finding additional information when needed. It is a useful way to think about onboarding an agent without expecting it to absorb the entire history of a business at once.
The phrase human in the loop comes up frequently in AI discussions, so I ask what it should mean in practice. Johnny describes checks before work begins and a review of test results and other evidence at the end, supported by progress updates along the way. The person responsible needs to set expectations and assess the outcome. Being available for an alert is only one part of that responsibility.
An unnamed streaming platform project provides a practical example. Johnny explains how Southworks developed the supporting process in stages, improving documentation, tests, and the infrastructure around the developer workflow. He reports a tenfold increase across requests, changes, and delivered value, although we do not have an independently verified measure or a detailed breakdown. His broader observation is that the same improvements made the environment easier for human engineers to understand and work in.
Finally, we discuss hiring and professional development. Johnny values software and cloud knowledge, curiosity, clear writing, time management, and the judgment to choose between competing possibilities. AI can make experimentation quicker, but someone still needs to decide which experiment deserves attention and when to stop.
I would love to hear how this compares with your own experience, so please share your thoughts. Where does work now wait in your engineering team, and how are you helping people make better decisions about what to build?