This article explains why AI in capital markets requires vertically deployed, workflow-native intelligence rather than generic chatbots. Written by Wilson Chan, Founder of Permutable AI it outlines how asynchronous, domain-specific AI creates real decision advantage in finance. It is aimed at investors, market leaders and institutions evaluating the future of AI-driven decision systems.
There’s a natural instinct in every organisation to start with an off-the-shelf chatbot. It’s accessible, it’s familiar, and it feels like progress. But when you move beyond experimentation and into mission-critical environments – finance, capital markets, commodities, risk – the limitations become obvious very quickly.
But in AI in capital markets, that approach breaks down quickly.
Capital markets are decision systems, not conversations
Traders, analysts and portfolio managers are not sitting in front of a blank chat window asking one question at a time. They are operating asynchronously, across multiple threads of information, under time pressure, with accountability. Decisions are rarely linear. They are iterative, probabilistic and contextual. A generic chatbot, no matter how advanced the model, is not designed for that reality.
This is the core misunderstanding we see in how AI in capital markets is being deployed today.
The real challenge in applying AI to capital markets isn’t model capability. It’s how outputs are structured, how context is preserved, and how intelligence fits into an existing decision workflow. That requires deep domain understanding – how desks operate, how risk is evaluated, how signals are weighed, and how decisions are actually made in practice.
Vertical deployment creates real differentiation
At Permutable AI, we’ve been deliberate about not competing with big tech on horizontal tooling. That’s not where our edge is. Our advantage is vertical: building AI systems that are native to specific market workflows. Systems designed around how capital markets professionals consume information, not how engineers think conversations should flow.
This is why we focus on asynchronous, multi-threaded intelligence rather than synchronous chat. Our users aren’t asking AI to “answer a question”. They’re using it to monitor narratives, assess risk, test assumptions and surface weak signals – continuously, in parallel, and in context. The interface matters because it shapes the quality of the decision.
Big technology platforms are optimised for scale across consumers and enterprises. They are not optimised for the nuances of a metals trading desk, a macro risk team, or an asset manager managing exposure across volatile global narratives. Nor should they be. That depth only comes from specialising.
Our conviction at Permutable AI
We believe the next phase of AI adoption in finance will be defined by embedded intelligence, not standalone tools. AI that lives inside the desk environment. AI that understands the language, incentives and constraints of its users. AI that augments judgement rather than attempting to replace it.
From an investor’s perspective, this is where durable value is created. Vertical AI systems benefit from higher switching costs, deeper data moats, and stronger alignment with revenue-generating workflows. They don’t win by being everything to everyone – they win by being indispensable to a specific user, in a specific context, making a specific class of decisions better.
That is the path we’re building towards at Permutable AI. Not a chatbot for markets – but an intelligence layer purpose-built for how markets actually work.
Built from market reality, not theoretical AI
Our perspective on AI in capital markets has been shaped by years of working directly with market participants who operate in environments where information is incomplete, narratives shift rapidly and decisions carry immediate consequences.
In practice, capital markets users rarely want “answers” in isolation. They want context: how today’s signal compares to yesterday’s, which narratives are strengthening or fading, and where consensus may be mispriced. They want to understand not only what is happening, but why it matters now and how it changes risk.
This is why our systems are designed around narrative evolution, multi-entity analysis and temporal awareness. Markets are not static datasets; they are living systems where meaning changes over time. Generic AI tools struggle here because they treat each interaction as discrete. Our approach to AI in capital markets treats intelligence as cumulative.
That design philosophy comes from deep engagement with real desks – commodities traders navigating geopolitical supply risk, asset managers balancing macro exposure, and strategy teams stress-testing assumptions under uncertainty. Each use case reinforces the same insight: AI must adapt to market behaviour, not ask markets to adapt to AI.
This lived experience is what allows us to build systems that feel intuitive to professionals, integrate naturally into decision workflows, and remain valuable long after the novelty of AI has worn off.
Partner with us to shape the future of AI in capital markets
If you’re building, investing in, or operating at the sharp edge of capital markets, we’d love to explore how we can work together. Permutable AI partners with institutions and forward-thinking organisations to deploy vertical, decision-grade AI where it matters most.
Get in touch with us at enquiries@permutable.ai to discuss partnership opportunities as we build the next generation of AI in capital markets.