From Vibe Coding to Vibe Analytics

This shift may send shivers up the spine of Data and Analytics professionals, where it is known all too well that trust in data is hard won, but very easily lost. However it is important to understand where this approach to working with data fits in as a complement to traditional BI and Analytics, rather than as a replacement.

The concept of ‘Vibe coding’ emerged from Andrej Karpathy’s influential X post about a new programming style where developers “fully give in to the vibes” and let large language models (LLMs) write code through natural conversation. In Karpathy’s words:

So what changed?

While “talk to data” solutions have existed for years, they never gained widespread adoption due to technical capability limitations. In short, they weren’t great. Earlier natural language processing (NLP) systems could only handle basic queries and often produced unreliable results. And while still far from perfect, today’s LLM GenAI capabilities have crossed a critical capability threshold, enabling complex analytical conversations that were impossible just a few years ago.

Consider an example: Instead of spending days constructing queries to analyse customer feedback, a business leader can now ask, “From our customer calls, what were the top themes for customers with lower NPS scores?” The system doesn’t just return a simple answer, it engages in a dialogue, surfacing unexpected patterns, suggesting follow up questions, and enabling real-time exploration of emerging insights.

If you have tried to use LLMs or any data specific ‘talk to data’ tools, you’ll know that it still has a way to go, and relies on really solid data foundations to be able to create trustworthy responses (think data quality, metadata, semantic and metrics layers, and well constrained data products), but the trend in the capability is undeniable.

The Fundamental Process Transformation

Traditional business intelligence follows a rigid, linear process that creates bottlenecks and often involves a number of handovers and dependencies between teams. The traditional flow involves someone with a business question working with a technical BI or Data Analyst, explaining their question or business problem (usually very poorly), their analyst interpreting this, going away and building queries and data visualisations, sharing this back and iterating on this cycle until their original question is answered.

This approach creates dependencies between those with business questions and those with the technical skills to be able to answer them, creating lost in translation problems and limits real unfettered exploration of data.

The improvisation aspect is crucial. Where traditional analytics asks static, constrained “What happened, why?” questions, vibe analytics allows for collaborative exploration to uncover both new answers as well as new questions. This combination of human intent and curiosity, with machine pattern recognition often reveals new insights that neither alone would discover.

Democratising Data Access

One of the most significant implications is the democratisation of analytical capabilities. Business users can increasingly explore and understand their own data without waiting for intermediaries to translate their questions into technical queries and data visualisations. This shift eliminates the frustrating hop between business needs and technical implementation.

The Evolution of BI and Data Analytics Roles

While this shift doesn’t eliminate the need for traditional BI and Data Analyst roles, it will transform some roles in two critical directions:

Second, some will evolve toward strategic advisory roles using the insights gained. Instead of spending time translating business questions into technical queries, they can focus on helping organisations understand what their new found answers mean and how to act on insights.

There is also the ongoing need for deep Data Analytics skills to validate data, assumptions and insights for high impact decisions and actions that may follow from discoveries made through.

Critical Success Factors

Michael Schrage shared the below caution in his MIT Sloan Management Review article: Vibe Analytics: Vibe Coding’s New Cousin Unlocks Insights:

• Data foundations – Ensure the underlying data is of high quality; good metadata is present including field level names with meaningful descriptions and synonyms if relevant; semantic and metric layers are used as required.
• Start small – Provide focused data with specific use cases in mind. Start with a small set of limited questions that you anticipate being asked, and then evolve and grow that as required.
• Provide good examples and instructions – When setting up the space for users to interact with, provide a limited, focused set of instructions including example sql queries where possible.
• Self test and iterate – Have a suite of test benchmark questions, and fine tune responses with specific feedback and feedback until you are getting reliable responses
• Test with a small group of users – Choose a small group of users to initially beta test with. Ensure they know their job is to help evaluate and train the model. Set expectations on the scope of the questions they should be focussing on and have them upvote, downvote and provide additional feedback on responses
• Monitor in production – Ensure ongoing ownership of model is clear and that performance against benchmarks are monitored. Make sure users are using common sense and know the limitations of this approach. If critical decisions are being made based on responses, ensure results are validated accordingly.

Where to from here?

Success requires both technical capability and cultural adaptation. Teams must learn to ask better questions, follow unexpected threads, and synthesise insights from this conversational exploration.

As we see capabilities continually improving with every model release, vibe analytics offers a glimpse into a future where data becomes a collaborative partner in discovery rather than a passive resource to be queried.

This shift mirrors the evolution from manual queries to automated monitoring systems, but with exponentially more sophistication. Agentic Analytics doesn’t just flag predetermined thresholds, it can dynamically discover what should be monitored based on evolving business contexts and emerging data patterns.

Beyond making people aware of actual or forecasted changes in data, agentic integration also allows for automated hypothesis testing as well a strategy formulation. Instead of a human having to define questions, create a hypothesis to explain causality, and then work to formulate a business strategy to react to the change, an agent might observe:

• Observation: Customer acquisition costs increased 13% in Q3, but this correlates with higher lifetime value.
• Hypothesis: We’re attracting more premium customers.
• Validation: Testing through cohort analysis and behavioural segmentation, calculating customer lifetime value trajectories with 90% confidence intervals to validate the premium customer thesis
• Recommendation: Allocate more resources to acquisition channels or campaigns that are driving premium customers, as higher lifetime value offsets the increased acquisition costs.
• Action: Run channel-level analysis to identify which platforms (e.g., paid search, social ads) are over-indexed for attracting high-value customers. Pilot A/B tests to validate strategy.

This blogpost was originally posted on Bryn’s LinkedIn in August 2025 and Bryn gave a Deep Dive session on this topic at AgileAus26. Images were sourced from the original LinkedIn post.

What am I actually built on? AI is now forcing that question on every knowledge worker, every leader, every organisation. And my answer to it matters more than it has at any previous point in my career.

Reinvention one: by instinct

My first reinvention was not strategic. It was instinctive. I was drawn to roles where I had to build things, solve problems that crossed functional lines, and learn fast. That pattern put me in the deep end in a range of roles across multiple countries, spanning multiple industries. Each context looked different on the surface, but shared something underneath.

One of my leaders saw this before I did. Early in my career, James Kim told me something I did not fully understand at the time. He said he could put me in any role, in any company, in any industry, and I would figure it out. It took me years to understand, believe and feel what he saw.

What transferred across every context was not domain knowledge. It was the ability to walk into unfamiliar territory, read the system, build the relationships, and deliver. The thing that moved with me was not what I knew. It was how I learned.

Reinvention two: by design

My second reinvention started with a tool and became something more fundamental. When I completed CliftonStrengths for the first time, it gave me language for what had been driving my career all along. My top talents were not personality traits. They were the operating system I had been running without being able to name it.

That clarity changed how I made decisions. But the sharper shift came from a conversation over a cup of tea.

Andrew Walduck, a leader I turned to often for career perspective, said something that has stayed with me since. “Hope is not a strategy, Nat.” He was not being unkind. He was being precise. He told me I needed to stop waiting for the right opportunity to find me and start being clear about three things: what I had done, what I excelled at, and what I wanted to do next.

So I wrote it down. I created a document I called “Hope is Not a Strategy”: a clear articulation of my strengths, my purpose and my values. I revisit it every year to update what I have done and what I want to do. When I can say with clarity that my purpose is to create more than I consume, and that where I add most value is building high-performance teams that can create an ambitious vision AND execute, I stop applying for roles and start designing my contribution.

Reinvention stopped being something that happened to me. It became something I practised. What made the difference was this: reinvention without self-awareness is just movement. You accumulate experiences rather than compounding them. When reinvention is anchored in purpose and strengths, each move builds on the last. You unlock a different kind of energy: one that comes from progressing your purpose, not just your career.

Reinvention three: by necessity

I have spent twenty years leading digital transformation. I thought I understood what it meant to lead through technological change. AI has humbled that assumption.

The pace is not just faster. It is qualitatively different. The technology I am using today did not exist six months ago. The technology I will be using in six months probably does not exist yet. The old model was to build expertise, rise through the ranks, and protect your position. It no longer holds. What you know is worth less each cycle. How you learn and adapt is worth more. Your current position is not a place to protect. It is a platform to launch from.

The risk I am most alert to is what I think of as the empty loaf problem.

Imagine a village where a machine arrives that can bake perfect sourdough in seconds. The baker, delighted, stops feeding his starter. He stops kneading by hand. The apprentice never learns to feel when the dough is ready. The bread looks the same. But there is nothing inside. When the machine breaks, there is no craft to return to.

The same risk exists when leaders delegate their thinking to AI. The output looks coherent. The communication sounds polished. But the underlying craft quietly atrophies: the ability to ask the question nobody else thought to ask, to push past the first coherent answer to the truer one beneath it, to synthesise complexity, to make a judgment call under pressure, to see what the data cannot yet confirm. You produce the loaf without cultivating the starter.

This is not an argument against using AI. I use it extensively but deliberately. The argument is about what you choose to protect.

The thread

If I zoom out across all three reinventions, the pattern is this: through each reinvention, what I had to let go were the things tied to a specific context: roles, titles, domain knowledge. What I held onto were the things that are most distinctively mine: the ability to learn anything and draw energy from that process; to take something broad and ambiguous and break it into something that can actually be executed; to understand what people need, not just what they ask for; to build relationships where real trust and real results coexist; to hold both the vision and the execution without losing sight of either. These are not just capabilities. They are my strengths: the expression of how I am wired to think, to lead, and to learn.

AI can augment every one of those strengths. It cannot replace them.

The difference between reinvention that compounds and reinvention that drifts is whether you know what you are building on. Your purpose. Your strengths. The things that do not change even when everything around you does.

AI is going to force this question on every knowledge worker: what are you actually built on? If the answer is your title, your technical skill or your current role, you are standing on shifting ground. If the answer is deeper than that, you have something solid to build from.

I am building again. It is uncomfortable. Again. And I am more certain than ever that this is exactly where I am supposed to be.

What are you protecting in this reinvention, and what are you prepared to let go?

This blogpost was originally posted on Natalie’s LinkedIn in March 2026.

Big thanks to guest blogger Caidyrn Roder who has shared his unique perspective alongside his paramedic co-speaker (and uncle!) Nick Roder, at AgileAus – not once, but twice.

Ambulance Victoria employs over 5,000 paramedics to provide emergency health care across the state. These paramedics are all trained to work outside of the hospital environment, which means that they have no defined workplace, limited managerial oversight and they are often remote from support.

And yet their performance is nothing short of remarkable. In 2023 – 2024, they attended over one million incidents. And 97% of patients surveyed reported that their experience was positive, and that they felt safe and cared for¹.

So when I learned about this, I got curious. How do they collaborate? How do they deal with the complexity? And what can we borrow from them to improve how we work in organisations?

In my opinion, the answer for many organisations is that they unintentionally work against their people.

I believe that we can learn from paramedicine, adapting the techniques that they use to create high-performing teams. Once we accept that people are our delivery system, we can work with them them by doing the following 4 things:

1. Design processes for humans
2. Create a just culture
3. Foster civil interaction
4. Flatten the authority gradient

Design processes for humans

Paramedics deal with cases that range from interhospital patient transfers to accidents to search and rescue. They have complex protocols and procedures which they have to perform under pressure. They work with people that they may have never met before. And all of this in challenging environments like accident scenes, remote areas or residences with access issues.

In response, paramedics have developed a range of tools that provide a systematic approach, constructed in a way that recognises that they will be deployed in a range of contexts while keeping the needs of patients and paramedics at the centre. In other words, they design processes for the humans who are involved.

Consider your organisation for a moment. I’d wager that you also have to deal with complexity. Do you have processes that don’t serve the people who use them? Or aren’t documented? Or are documented but not followed? The insight that I think we should take from paramedicine is that processes that are not designed for humans, will not be used by humans.

The good news is that there are simple steps that you can take to design processes for the humans in your organisation.

• Listen to your people – Talk to your teams. Talk to your people leaders. Talk to your executives. Listen carefully to understand the current context, what’s working well and where there are opportunities to improve.
• Involve the people who do the work – The people who understand the work and the environment are the best ones to (re)design the processes. So make sure that you involve them. Better yet, empower and support them to continuously improve the processes themselves.
• Prioritise simplicity – When you’re designing process changes, challenge yourself to make them as simple as possible. Could you have fewer steps? Do you really need to involve all of those people in a decision? When faced with a choice, prioritise creating a simple process that covers 80% of cases over a complex one that covers 100%.
• Continuously improve – I hate to be the one to tell you this, but you won’t get your processes right the first time (or even second, third or fourth). So build in continuous improvement from the start by creating an environment that encourages your people to continuously try new ways of collaborating.

Create a just culture

Continuous improvement is not just important for paramedics, it’s non-negotiable. For them, raising concerns can lead to valuable changes in practice that improve outcomes for patients and paramedics. This is only possible in an organisation that both makes it safe for employees to speak up, and is prepared to listen.

In other words, continuous improvement is about much more than just processes and tools. You can’t simply tell teams to do retrospectives or to implement tools like lean A3s and leave it at that. To unlock the true power of continuous improvement you need to create a culture where people are comfortable talking about and learning from their mistakes. A culture where people are empowered to change the way they work and to challenge existing practices.

Paramedics call this a just culture. A just culture means that people are able to raise concerns, mistakes or challenges within the organisation in an environment that treats employees with due respect. A just culture recognises that human error is inevitable and that complex systems designed by humans will fail². There is shared accountability between the organisation (the system) and the individual when things go wrong.

In a just culture, you don’t blame or punish individuals when their actions or decisions are commensurate with their experience or training. Instead seek to understand the broader context for their actions so that you can identify improvements. This is not the same as a no-blame culture. People are still held accountable for their actions, and negligence and willful misconduct are not tolerated.

There are 3 simple things that you can do to foster a Just Culture in your organisation³.

1. Support the people involved – When something goes wrong, your first priority should be to support the people involved. Making a mistake is rarely a positive experience, however a supported employee is more likely to resolve their experience productively and learn and grow as a result. They are also more likely to talk about mistakes in the future.
2. Manage your biases – Give people the benefit of the doubt by assuming that they set out to do their best and achieve a good outcome. When someone makes a mistake, put yourself in their shoes to ensure that you are empathetic to them. Try asking yourself this question: Could another person with similar knowledge and experience make the same decision in the same situation?
3. Consider the system – Rather than focusing solely on individuals and their actions, you need to recognise that they are working within a broader context. Remember, a Just Culture recognises that there is shared accountability between the organisation (the system) and the individual when things go wrong. So ask: What aspects of the system contributed to the actions they took, and the decisions they made? Why did it make sense to them at the time?

Foster civil interaction

There is a growing body of evidence in both medicine and business that rudeness and incivility undermine team effectiveness. In their 2015 study, Riskin et al concluded that “not only does rudeness harm the diagnostic and procedural performance of [medical] practitioners, it also seems to adversely affect the very collaborative processes that might otherwise allow for teams to compensate for these effects”⁴. In other words, incivility affects both an individual’s cognitive abilities as well as their ability and willingness to collaborate with and help others.

Incivility includes overt behaviours like mocking, teasing or insulting. Or yelling, slamming doors, or talking about people behind their backs. It can also be more subtle behaviours like responding to emails in a meeting, not smiling or acknowledging others when you pass them in the hall or not listening when someone is speaking to you.

Importantly, perceptions of incivility can vary from person to person. What one person finds rude, might be totally acceptable to another person.

Incivility is a common cause of stress in the workplace. For example, given that paramedics often work in traumatic circumstances you might expect their main cause of stress to be the work itself. However, they often report that how people behave has a larger impact on their stress levels.

Rudeness can also lead to reduced creativity, lower productivity, increased absenteeism, lower quality of work and in the worst case, employee churn and damaged customer relationships.

Interestingly, the impact is not just on the person on the receiving end of the behaviour. People observing uncivil behaviour are also impacted. So, if a team lead is rude to a team member in a meeting, the whole team’s performance may suffer as a result.

The implications for organisations are profound. Fostering civil interaction not only supports staff wellbeing, but it can also improve team effectiveness. So what can you do to foster civil interaction in your organisation?

• Model good behaviour – Be conscious of how you interact with your colleagues and hold yourself to a high standard. Avoid mocking, teasing or insulting people. Even if you think that it is “good natured” it’s possible, even likely, that others don’t see it that way.
• Ask for feedback – Feedback is a good way of monitoring how you’re doing. It helps you improve by giving you other perspectives, and it builds psychological safety and encourages a growth mindset in those around you.
• Establish team or company norms – Norms are an agreed set of behaviours that are acceptable in your context. They help you be clear about what behaviour is expected and what won’t be tolerated.
• Take bad behaviour seriously – Organisations who don’t address rude behaviour often pay the price. When leaders ignore bad behaviour, they are giving their implicit approval of the behaviour. Staff stop reporting incidents because they feel that nothing will be done. The behaviour continues unabated, stress levels increase and often the simplest solution for staff is to quit.
• Include civility in recruitment and employee training practices. Look out for indications of rude behaviour in interviews. Get multiple perspectives by involving team members in the interview process. Make people aware of the concepts and the impact in training

Manage your authority gradient

Authority gradient is a concept borrowed from the aviation industry and adopted by medicine and paramedicine. It refers to the established, or perceived, command and decision-making power hierarchy in a team or organisation⁵. It is the status difference (real or perceived) between 2 or more people in a team.

At one end of the spectrum, an extremely steep authority gradient is where an individual holds all of the power. An example is a team lead who has a controlling or prescriptive management style – “it’s my way or the highway”. Team members are expected to do what they’re told and not ask questions or express opinions. This results in fewer perspectives and ideas, reduced communication and collaboration, lower team engagement and an environment where feedback is not given freely.

At the other end, an extremely shallow authority gradient is where everyone’s opinion is valued equally on all topics. There is no difference between experienced senior team members and more junior ones. Everyone’s voice is heard and feedback is given freely, but decision-making may be slow and/or poor and responsibilities may be unclear.

Studies show that getting authority gradient wrong can have disastrous results. I’ve seen the impact on teams and organisations myself. I once worked with a team developing a proof of concept for a new digital product with the primary purpose of learning what users wanted. The cross-functional team was talented and experienced, but they were hamstrung by a leader who had an authoritarian leadership style and a fixed view of how things should be done. The team did not feel safe to suggest a better approach, so they followed the leaders directions slavishly. The result was a lot of wasted money and effort and very little learning.

Authority gradient is also particularly important for work on complex problems (i.e. most of the work we do in organisations), because this entails a high degree of uncertainty and risk. In organisations we often have to make decisions based on incomplete information, under conditions of high uncertainty and risk, where there is no obvious right or wrong answer, where people hold conflicting opinions and decisions and actions are only proven right or wrong by the results they achieve.

Interestingly, the concept of managing an authority gradient challenges traditional views on leadership. All paramedics can be both leaders and followers, adjusting dynamically depending on the situation. For example, it’s not uncommon for the first crew on an accident scene to retain the leadership role, even when a more senior paramedic arrives. They may even enlist that more senior person as a ‘follower’, leveraging their skillset to get the best outcomes for their patient.

Being a follower does not mean passively following orders. In high-performing teams, followers anticipate the next step, offer different perspectives and solutions and demonstrate initiative.

There is no single “correct” authority gradient for a team or organisation. It changes based on circumstance. The most effective teams are conscious of their authority gradient and have protocols in place to adapt it dynamically. So while there is no one-size-fits-all approach, there are some heuristics that you can use to determine what is appropriate for the circumstance.

• Shallow authority gradient – Well-suited to dealing with complex problems. Creates a space where team members feel safe to contribute ideas, ask questions, challenge each other and debate solutions. A shallow gradient also helps to create a learning environment because people are more likely to give feedback and share learnings when things go wrong.
• Steep gradient – Appropriate in times of crisis, when fast decisions made by an experienced leader are more important than getting a range of perspectives. Be careful not to maintain a steep gradient for extended periods of time as this often makes teams less productive and reduces learning, morale and engagement.

Conclusion

Paramedics consistently achieve remarkable results in high-pressure, complex environments because they understand that people are their delivery system. Organisations seeking to emulate this level of performance can draw profound lessons from the field of paramedicine. By designing processes for humans, organisations can empower their workforce to operate efficiently even in complex scenarios. Furthermore, cultivating a just culture that not only tolerates but actively encourages learning from mistakes is paramount. This creates an environment where individuals feel safe to speak up, share insights from failures, and continuously improve, rather than hiding errors due to fear of blame. Promoting civil interaction within teams is also crucial, as incivility erodes trust, impairs cognitive function, and undermines collaboration. Finally, consciously managing the authority gradient – adapting it dynamically to the demands of the situation – allows for optimal decision-making and leverages the diverse expertise within a team.

Ultimately, investing in these human-centric principles will not only improve internal processes; it will unlock the full potential of your people.

Stay tuned for more information on AgileAus26 – we cannot wait to gather with curious minds and rethink the way we work and live!