The financial services industry is in the middle of an infrastructure revolution. Trading floors that once ran on fixed turrets and on-premise recorders are migrating to cloud-based voice systems.
We sat down with Matthew Carey at Theta Lake to talk through what is driving this shift, where legacy tools fall short, and what firms need to get right before they go live on new infrastructure.
Legacy Tools Were Not Built for a Distributed World
For decades, voice compliance was straightforward. You knew where the phone was. You knew where the recorder was. The recording estate had a clear perimeter.
Cloud changes all of that. Communications are distributed across geographies, devices, and platforms. The pace of change: new feature releases, new integrations, new modalities, is far faster than on-premise systems were ever designed to absorb.
“Some of the legacy applications start to not so much fail, but struggle to keep pace with the new changing environments we see from a cloud perspective,” Carey explains.
The richness of metadata and content increases in cloud environments, purpose-built surveillance tools become increasingly important to preserve context as communications move across modalities and platforms.
What’s Driving the Migration
The push toward cloud voice is not a single force, it is several converging pressures.
A significant portion of voice infrastructure in financial services is sitting on aging on-premise equipment approaching end of life, with vendors already setting those dates. At the same time, the business is asking for more: traders no longer want their turret limited to a device on a desk. They want mobile access. They want the ability to work from anywhere without losing their core voice capabilities.
On-premise environments simply do not offer the interoperability to support that. Historically, on-prem systems connected to a recorder and maybe a CRM tool. Today, firms expect communications to integrate with the full stack: Teams, Zoom, Bloomberg, and the other workflows that live on a trader’s desktop. That level of integration is a cloud-native story.
The Multimodal Compliance Gap
Modern enterprise communication does not stay in one channel. A single business interaction might begin as a voice call, move into a Teams chat, continue in a video meeting, get summarized by an AI assistant, and close out over email – all within minutes.
From the user’s perspective, that is one conversation. From a compliance team’s perspective, it can look like five separate, disconnected interactions spread across five different platforms.
“Without applying a holistic framework across all of those modalities, you’re at risk of losing the full context of the conversation,” Carey notes.
The surveillance challenge is not just capturing each piece, it is stitching them back together in a way that gives reviewers the same context the participants had. That means ingesting from each platform, reconstructing the thread as it actually happened, and making it searchable from a single location rather than requiring compliance teams to hunt across systems.
Trader Voice: The Compliance Laggard
Trader voice occupies a paradoxical position in the compliance landscape.
It was among the earliest communication channels to be regulated, with recording obligations significantly expanded under Dodd-Frank and similar global regulations. Yet despite decades of mandated recording, it remains one of the least mature when it comes to surveillance capabilities.
“It’s probably the communication tool that lags in maturity when it comes to surveillance,” Carey says.
The reason is structural: the majority of trader voice infrastructure is still on premise, captured by dedicated recorders, and has not moved at the same pace as other communication tools.
That is changing. The major trader voice providers like IPC, Cloud9, and BT are all advancing their cloud offerings, and there is growing demand from traders for mobile access and deeper integration with platforms like Microsoft Teams. Theta Lake’s recently announced partnership with IPC reflects this shift, aimed at helping firms bring their trading communications and surveillance capabilities up to the same standard as their electronic communications today.
The Real Cost of Going Live Without Compliance Architecture
Under MiFID II, Dodd-Frank, and related regulations, financial firms have strict recording obligations for trader communications. When new infrastructure goes live before compliance architecture is addressed, the result is a system built to be bolted onto; and bolt-ons have real consequences.
“By bolting it on after the fact, you’re ultimately going to hamper surveillance capabilities downstream,” Carey explains.
Each new modality like mobile SMS, AI-generated meeting summaries, chat threads; that gets added after initial deployment creates another layer of fragmentation for compliance teams to manage.
The complexity compounds quickly. It is no longer just about capturing what traders say; increasingly, it is also about capturing what AI systems say on their behalf. AI-generated summaries, transcriptions, and analysis are becoming part of the conversation. A unified capture framework that brings all modalities under the same supervision umbrella from the start is far more effective than assembling it after the fact.
What Purpose-Built Cloud Compliance Actually Looks Like
When Carey describes what firms should evaluate in a cloud-native compliance solution, he keeps returning to one word: context.
The goal is not just to capture communications, it is to preserve them in a way that gives surveillance teams the same contextual view that the participants had. That means being able to replay conversations as they occurred, across every modality they spanned, from a single interface.
Looking ahead three to five years, Carey sees the focus shifting toward what Theta Lake calls the communication lifecycle: a framework for proving, at every step, that communications were captured correctly from source to archive.
“Being able to prove that settings were as they should be as that communication moved from the source system, like MS Teams, to the archive within Theta Lake – making sure the integration components are set correctly, making sure the user is set correctly for archiving, and being able to present all of that back to regulators and internal stakeholders.”
That kind of end-to-end assurance is where voice compliance is heading. And for firms still navigating a legacy-to-cloud transition, the lesson is clear: the earlier compliance is part of the conversation, the better the outcome.
Copyright © 2026 RegTech Analyst
Copyright © 2018 RegTech Analyst
