Evolving essays, not a blog.

There is a question we have learned to ask about any technical field, because the answer reveals almost everything about its maturity:

Are the people who build the thing the same people who use it?

In young fields, the answer is always yes. The Wright brothers built their aircraft and flew it. Early motorists were, of necessity, mechanics; the first computers were programmed by the engineers who wired them. When a technology is new, construction and use are a single craft, held in a single pair of hands, because no one else can hold it.

And in every field that matured, the answer became no. Aviation split into aerospace engineers, pilots, and passengers — three professions, or a profession, a skill, and no skill at all, connected by machines and institutions engineered to make the separation safe. Medicine split into the people who design molecules and the people who prescribe them. Computing split so many times we stopped counting: chip designers, operating system engineers, application developers, and finally users, billions of them, who have never seen a line of code and never need to.

This separation is not a compromise that mature fields reluctantly accept. It is the mechanism of their maturity. The moment users stopped needing to be builders, each discipline exploded — in scale, in safety, in usefulness, and, not least, in the sophistication demanded of its builders. Nobody thinks aviation was diminished when passengers stopped stitching wing fabric.

Energy system modeling has not made this transition. Forty years in, it remains a field where the person who builds the model is expected to run it, maintain it, interpret it, and defend it — and where anyone who wants answers from a model is expected, implicitly, to first become the kind of person who builds one. We are still stitching our own wing fabric. And then we hold conferences to ask why so few people fly.

The artisanal trap

It is worth being precise about how the field got here, because nobody chose it.

Energy system models were born in research institutions, and they took the shape of the institutions that made them. A model was a dissertation: one country, one researcher, several years. The India MARKAL model — the first comprehensive energy system model of that country, and the beginning of our own story — was built exactly this way in 1990, as doctoral work, with database tools that would now look archaeological. That was the only way such a thing could be built at the time, and it was a legitimate scientific achievement.

But the dissertation-shaped model created a dissertation-shaped field, and the shape hardened into assumptions we stopped noticing: that a model is inseparable from its maker — so when the PhD student leaves, the model dies. That building is the qualification for using — so the community's entire pedagogy teaches construction, and almost none of it teaches interrogation. That every new question deserves a new model — so the field's effort pours into building the same power sectors, the same technology databases, the same base years, over and over, in parallel, forever. Energy modeling has spent decades hand-building aircraft one at a time, each aerodrome incompatible with the next.

The costs are the field's most familiar complaints, though they are rarely traced to their common root. The modeler bottleneck: the pool of people who can build serious models is tiny, so analysis queues behind a scarce craft skill. The black-box accusation: decision-makers distrust results they had no hand in shaping, because the only mode of participation on offer — become a builder — is one they cannot accept. The graveyard of models: enormous public investments that produced sophisticated models exactly once, for exactly one study, and were never run again. These are not separate diseases. They are all symptoms of one condition: the field has not yet separated its builders from its users.

The objection that deserves an answer

The standard defense of the status quo is serious and deserves to be met head-on: you cannot responsibly use a model you don't understand, and you cannot understand a model you didn't build.

The first half is true. The second half confuses two different kinds of understanding — and every evolved domain has learned to tell them apart.

A pilot understands an aircraft. Not as its builder does — she could not design a turbine blade — but operationally, deeply, and in exactly the dimensions that flying requires: its envelope, its failure modes, its instruments, its behavior at the edges. Her understanding is not a lesser version of the engineer's. It is a different profession's understanding, built through different training, certified through different institutions, and it is the understanding that actually lands the plane.

What a model user owes the model is the same: not knowledge of its construction, but command of its behavior. What drives this scenario? How does the answer move when assumptions move? Where is the model strong, and where does it thin out? Those questions can be taught, and answered, without a single line of GAMS — provided the model and its tools are built to expose themselves.

That proviso is the real requirement, and it is why the separation of builders and users cannot be achieved by declaration. It has to be engineered. Users need three things that builders were able to live without: transparency — every number traceable to its source and its logic, because users cannot read trust out of the code; participation — the ability to change assumptions and interrogate outcomes directly, because ownership of insights comes only from handling them; and reproducibility — the same model, the same data, the same answer, for everyone, every time.

Notice what is not on that list: source code. The open-source movement, for all it has given science, quietly assumes the builder-user identity — it offers participation only in the builder's medium. Publishing half a million lines of model generation code is a gift to the seven people on earth positioned to read it, and an alibi to everyone else. We have come to call the alternative Open Use: closed workshop, open vehicle; the meal delivered, not the recipe recited. Judge the model by whether you can drive it, inspect it, and reproduce it — not by whether you could, in principle, have welded it yourself.

There is also a newer reason the objection is losing force, and it would be dishonest not to name it. For forty years, construction was the bottleneck, so construction skill was the field's scarcest and most celebrated asset. That era is ending in front of us. Automation — and now AI — is making building cheap. What it cannot make cheap is judgment: knowing which questions to ask, which assumptions matter, which results to believe. When building is abundant and judgment is scarce, a field that trains only builders is training for the wrong scarcity. The user — the professional interrogator of models — is about to become the more consequential role. We should be building that profession deliberately, not waiting to improvise it.

Thirty years of performing the separation

We did not arrive at this argument by philosophy. We arrived at it by doing the work, in stages, each stage teaching us the next — which is why we hold the view with some confidence: we have watched the separation succeed every time we widened it.

First, we elevated building — and this remains our core competence. Veda began in the early 2000s as a results tool for the first global models and grew into the environment in which TIMES models everywhere are built and run. Its purpose, from the beginning, was to take the mechanics of modeling — data handling, scenario management, the thousand clerical cruelties of a large model — away from the modeler, so that a builder's time went into judgment instead of bookkeeping. This is the one thing KanORS has always done and never stopped doing: raising the cognitive level at which building happens, so that the builder's scarce hours are spent thinking about energy systems rather than shepherding spreadsheets. Veda made builders faster and their work deeper. But it did something more important still: by standardizing how models are assembled, it made models legible beyond their makers. It was the assembly line that made every later step thinkable.

Then, we created users — and watched it work. KiNESYS was the wager that decision-makers could drive models they did not build: bespoke, high-resolution models, delivered as environments to explore rather than reports to receive. The wager paid off in the field's most demanding settings — World Bank climate and development strategies, national decarbonization pathways, the in-house modeling teams that global consultancies now run on the platform under our guidance. The people exploring these models are not TIMES builders. They are the field's first generation of professional users, and their existence is the empirical answer to whether the separation is possible.

Now, we are making users of everyone — and giving builders a head start no builder has had. KiNESYS proved the separation works when someone pays for the car. VerveStacks removes the last condition. Because three decades of construction knowledge could be compressed into an automated pipeline — global datasets integrated, reconciled, spatially and temporally structured, validated, documented — a professional-grade model of nearly any country on earth can now be generated in minutes. Which means it can be given away. Ready-to-run models, free, for every researcher, analyst, and ministry that was ever told to come back after learning to build.

And note who else these models serve. Every VerveStacks model ships as a transparent, fully documented, extensible structure — nearly complete national datasets with every parameter traceable to its source. To a user, that is a vehicle. To a builder, it is a foundation: the months of data assembly that once preceded every real question, already done, ready to be extended, enriched, and made one's own. The same artifact serves both professions, because it was built to the standard both require. What began in 1990 as one country, one researcher, and years of work is now every country, any user, and an afternoon — or any builder, and a running start.

That trajectory — from artisanal build, to industrialized building, to professional users, to universal users — is not our product roadmap dressed up as history. It is one experiment, run for thirty years, on a single question: what happens to energy modeling when you separate the builders from the users? The answer, at every stage, has been: it gets bigger, faster, more trusted, and more useful.

What builders should hear in this

If you have spent a career building models, this essay may read as a demotion. It is the opposite, and both the evolved domains and our own history say so plainly.

Start with our history, because it settles the question of allegiance. KanORS is, at its core, a builders' company. The thing we have actually done for thirty years — the competence everything else rests on — is making building a higher-order activity: absorbing the mechanical into tools so that what remains in the builder's hands is design, structure, and judgment. Veda was built for builders. VerveStacks' pipeline is builder's knowledge, industrialized. Even our models for users exist only because we held building to a standard — transparency, richness, reproducibility — that made them usable beyond their makers. The user profession we argue for is not a turn away from builders. It is what taking building seriously produces, in the end, in every field.

Now the evolved domains. Aerospace engineers did not lose standing when passengers stopped building planes; they became one of the most specialized and respected professions on earth, precisely because a world full of passengers demands far better aircraft than a world of hobbyists ever did. The separation did not shrink the builder's craft. It raised its stakes.

The same is coming here. A field with ten thousand model users will need better builders than a field with three hundred builder-users — builders held to industrial standards of validation, transparency, and reproducibility, building models that must survive contact with users who cannot silently patch their flaws. Building will get harder, more consequential, and more valued. What ends is not the builder's profession but the builder's monopoly — the era when the only door into modeling was the workshop door, and the field's reach was capped at the number of people willing to apprentice there.

Every discipline that matured has faced this moment, and none has regretted it. The question for energy modeling is no longer whether the separation will happen. Between the automation of construction and the arrival of AI, it is happening. The question is whether it happens by design — with the transparency, the training, and the institutions a real user profession deserves — or by accident, badly.

We have spent thirty years on the design. The tools exist. The models exist. The first users exist. And the builders — better equipped than they have ever been — remain at the center of it, because a field of users runs on the work of builders.

Both professions are open. Come build. Come drive.

What Does Democratization Mean?

Democratizing energy modeling involves making models accessible to a broader audience, removing barriers that prevent people from engaging with them. This includes:

Empowering Modelers Without Coding Expertise: Not every energy professional has programming skills, nor should they need to. Democratization means providing tools that allow modelers to focus on building scenarios and interpreting results without being bogged down by technical mechanics.

Engaging Stakeholders in Decision-Making: Policymakers, business leaders, and other decision-makers often rely on model outputs but rarely interact with the models themselves. Democratization ensures that these stakeholders can explore scenarios, understand assumptions, and develop ownership of insights.

Fostering Cross-Disciplinary Collaboration: Energy transitions involve complex systems that intersect engineering, economics, policy, and more. Democratized tools break down silos, enabling people from different disciplines to collaborate meaningfully.

Why Democratization Matters

Bringing Diverse Perspectives to the Table: Modelers excel at building and refining models, but stakeholders bring vital real-world context. For example: Policymakers know the intricacies of regulations and priorities. Business leaders understand market dynamics and investment strategies. Environmental advocates highlight equity and sustainability concerns. Democratized tools allow these perspectives to shape models, ensuring outputs are grounded in reality and aligned with diverse needs.

Enhancing Ownership of Insights: Insights are more likely to drive action when stakeholders feel a sense of ownership. If decision-makers can interact with models—exploring assumptions, testing scenarios, and interpreting results—they are more likely to trust the insights and implement them effectively.

Accelerating the Transition to Clean Energy: The energy transition is time-sensitive. Traditional modeling approaches, often confined to a limited group of technical experts, slow down progress. By enabling more people to engage with and benefit from models, democratization accelerates decision-making and implementation of decarbonization strategies.

Overcoming Bottlenecks in Expertise: The pool of modelers with advanced technical skills is limited, creating bottlenecks in the modeling process. Democratized tools empower energy professionals who lack programming expertise to contribute, expanding the modeling workforce and increasing overall capacity.

Driving Innovation Through Collaboration: Fresh perspectives often lead to breakthroughs. Non-modelers—such as domain experts, policymakers, and industry leaders—can challenge assumptions and propose novel approaches. This diversity of thought pushes the boundaries of what energy models can achieve.

How KanORS Champions Democratization

At KanORS, we are committed to making energy modeling more inclusive and accessible. Our tools are specifically designed to break down technical barriers and foster collaboration:

Veda: Veda is a domain-aware platform that allows energy professionals to build and use TIMES models without requiring programming expertise. It simplifies data management, scenario building, and result analysis, enabling more users to engage with modeling effectively.

KiNESYS: KiNESYS offers bespoke, ready-to-use models tailored to stakeholders. Decision-makers can explore insights, test scenarios, and understand implications without needing technical knowledge. This shift from building models to interacting with insights empowers users to focus on strategy and implementation.

The Path Ahead: A Call for Inclusivity

Democratizing energy modeling is about more than just tools—it’s about fostering a cultural shift in how we approach energy transitions. By making modeling accessible to a wider audience, we unlock the full potential of collaboration. Stakeholders from diverse disciplines can contribute their expertise, ensuring that insights are comprehensive and actionable.

As climate challenges grow more urgent, democratization ensures that energy models are not just tools for a select few but resources for everyone invested in a sustainable future. At KanORS, we are leading this movement, creating tools and frameworks that enable meaningful engagement and collaboration.

The Challenge of Silos in Energy Modeling

Technical Silos: Modelers, by necessity, focus on building and refining models. However, this technical focus can lead to disconnects between modelers and those who use or benefit from the insights, such as policymakers, engineers, and economists. This gap often results in models that are too complex for non-experts to engage with effectively.

Sectoral Silos: Different sectors of the energy system—electricity, transportation, industry—are often analyzed independently. However, real-world systems are deeply interconnected. A shift in the transportation sector (e.g., the adoption of electric vehicles) directly impacts electricity demand, while decarbonizing industry influences resource allocation across multiple sectors.

Geographical Silos: National and regional models frequently overlook global interlinkages like energy trade, cross-border energy infrastructure, and international climate policies. Without accounting for these dynamics, insights can be incomplete or misleading.

Why Collaboration Across Silos Matters

Integrating Diverse Expertise: Modelers bring technical precision, engineers add operational insights, economists provide market dynamics, and policymakers contribute context and priorities. By working together, these perspectives enrich the modeling process and produce outputs that are both comprehensive and actionable.

Breaking Down Barriers to Implementation: The ultimate goal of modeling is to drive real-world action. If stakeholders cannot engage with models due to complexity or lack of alignment with their needs, the potential impact diminishes. Collaboration ensures that models remain relevant and accessible to those who rely on them.

Capturing Interdependencies: Energy systems are deeply interconnected, and decisions in one sector or region often have ripple effects elsewhere. Collaboration across silos allows models to capture these interdependencies, leading to a more holistic understanding of the energy system.

Driving Innovation: Collaboration fosters creativity. When people from different disciplines and backgrounds come together, they challenge assumptions, inspire fresh ideas, and push the boundaries of what is possible.

How KanORS Bridges Silos

At KanORS, collaboration is embedded in our approach to modeling. Our tools are designed to foster engagement across disciplines, sectors, and geographies:

Veda: Simplifies technical workflows, enabling modelers to collaborate effectively with non-programmers and domain experts.

KiNESYS: Provides stakeholders with tailored models that allow them to engage with insights directly, bridging the gap between technical modeling and decision-making.

Sectoral and Regional Integration: Our frameworks capture interdependencies across industries and consider global energy trade, ensuring comprehensive and actionable insights.

Lessons Learned from Collaboration

Through our work, we’ve seen how collaboration across silos drives progress:

New Insights Emerge: Diverse perspectives uncover solutions that a single discipline might overlook.

Engagement Drives Action: When stakeholders feel ownership of insights, they are more likely to implement solutions effectively.

The State-of-the-Art Evolves: Collaborative processes advance methodologies and tools, pushing the boundaries of what energy modeling can achieve.

A Call to Collaborate

The energy challenges of today demand more than isolated expertise. They require a collaborative, cross-disciplinary approach that integrates technical precision, sectoral understanding, and global perspectives. At KanORS, we are committed to fostering these connections and enabling energy modeling to serve as a catalyst for meaningful action.