The conversations in Amsterdam this year cut through the noise. Forget the hype around LLMs, headline AI projects, and generic automation slogans. The real transformation in fixed income is happening where nobody was looking, in chat.
For decades, chat has been the beating heart of the OTC markets. Trillions in RFQs, prices, and orders negotiated daily in chat, only to vanish into the void at the end of the day. Unstructured. Uncaptured. Untapped. This year, that story flipped. The combination of open chat APIs and mature AI tooling has turned chat from a passive communication layer into the new automation surface for capital markets.
There’s no polite way to put it: trading desks are still crippled by manual workflows.
Copy-pasting prices. Re-entering orders. Reconciling trades from chat and spreadsheets into booking systems. The entire ecosystem runs on digital duct tape.
But 2025 marks the first time we’re seeing the fix at scale. With Bloomberg opening IB chat APIs - catching up with similar moves from other platforms - data and workflow can finally move directly from IB chat into the systems that matter. The workflows that once lived in traders’ heads are becoming executable, digital, and auditable.
The shift looks like this:
Unstructured Chat ➝ Structured Information ➝ Automation ➝ Control
The firms leading the charge are collapsing data silos, automating the repetitive, and building compliance into the process itself. This isn’t just about digitisation, it’s about orchestration too.
The buy vs. build debate around AI surfaced again at FILS, but this time, it felt settled. The majority of in-house AI builds in financial markets still stall before production. The reasons are predictable: lack of domain expertise, missing orchestration layers, disconnected data, and compliance headaches.
The consensus among practitioners was clear: AI is becoming a utility.
You don’t build your own grid; you plug into it.
You don’t build your own data centre; you use the cloud.
You don’t build AI infrastructure; you connect to one that’s already secure, compliant, and production-ready.
Firms that treat AI as an ecosystem, not a lab experiment, are the ones already extracting value. They’re augmenting workflows, not replacing people.
This was the most forward-looking theme at FILS, the move from static bots to intelligent agents. The step-change is intent recognition. Instead of waiting for a human prompt (“show me live runs”), agents act contextually (“surface matching liquidity across desks”).
They consolidate data across rooms, trigger risk checks mid-conversation, and execute workflows from chat without switching applications.
They don’t eliminate humans; they amplify them.
The future isn’t human versus machine. It’s networked collaboration: human to agent, agent to agent, all within guardrails that enforce compliance, permissions, and auditability.
The automation race isn’t just about speed; it’s about control.
Real-time permissioning. Complete audit trails. Embedded compliance.
The firms getting this right are not just faster; they’re safer, more transparent, and more scalable. In an era where regulators and counterparties are scrutinising data lineage and model explainability, control has become a competitive advantage.
The mood in Amsterdam was pragmatic. Leaders are done with pilots. They want production workflows, measurable ROI, and resilience under pressure. The winners will be those who can take the “digital exhaust” of chat, structure it, automate it, and turn it into decision-ready intelligence.
That’s the transformation ipushpull is driving, connecting chat, data, and AI into a unified automation layer across front-office workflows.
Chat isn’t just where markets talk. It’s where they now transact, automate, and control.
The next era of fixed income efficiency won’t come from replacing systems. It will come from rethinking the conversations that run them.
Please get in touch if you want to learn about what you can do with your real-time chat data for capturing and booking trades from chat, chat data mining and using quotehub to manage axes, runs, RFQs and quotes.