BMLL Technologies, the independent provider of harmonised historical order book data, and Tradefeedr, the network-based trading analytics platform, have announced a partnership to extend Tradefeedr’s analytics capabilities into equities and futures. The initiative, which includes a year-long industry pilot, represents Tradefeedr’s first move beyond its established FX analytics franchise and signals a broader ambition to become a multi-asset analytics layer for AI-driven trading workflows.
The partnership pairs BMLL’s Level 3, 2 and 1 historical datasets across global equities, ETFs, futures and US equity options with Tradefeedr’s analytics APIs and its connected network of more than 100 institutional clients. The combined offering will deliver enriched, standardised trading data through a single API designed, both firms say, to serve as the foundational data layer for a new generation of LLM-powered execution analytics.Defining “AI-ready”
The announcement leans heavily on the concept of “AI-ready” analytics, a term that invites scrutiny. In practice, Tradefeedr has built an agentic interface that sits on top of its analytics APIs, allowing clients to query their trading data using natural language. The system identifies the relevant APIs, executes the query, and returns structured results, effectively replacing the traditional dashboard-and-PDF-report model that dominates institutional TCA today.
In conversation with TradingTech Insight, Balraj Bassi, CEO of Tradefeedr, is blunt about the competitive implications. “All your traditional dashboards, costly interfaces, reports and traditional TCA providers are going to be in a bit of trouble,” he says, “because you can replicate what they do and build your own.”
The underlying architecture is LLM-agnostic – clients can choose between Claude, OpenAI or Google models – with the differentiation residing in the orchestration layer: the agent workflows, data governance controls, vector databases and MCP integrations that sit between the language model and the trading data. Tradefeedr has built a library of pre-configured workflows covering common analytics tasks such as algorithm performance evaluation, broker reviews and cost estimation.
The data gap
For Tradefeedr, the expansion into equities and futures has been client-driven. Its advisory group has been requesting multi-asset coverage for some time, but the firm lacked access to the quality of historical market data needed to underpin credible benchmarking outside FX.
Paul Humphrey, CEO of BMLL, frames the partnership in those terms. “The only barrier was not Tradefeedr’s capabilities,” he says. “It was getting their hands on the best quality market data, and that’s where this partnership comes in.”
The partnership is enabled through Tradefeedr’s participation in BMLL’s Activate: Data Credits Programme, which provides qualified partners with subsidised access to BMLL datasets to build and validate new products, with a route to long-term commercial deployment. For BMLL, which was acquired by Nordic Capital in October 2025, the arrangement offers a distribution channel into front-office analytics workflows without having to build its own client-facing product.From FX to multi-asset
An obvious question is whether Tradefeedr’s FX-native model translates to the different microstructure dynamics of equity and futures markets, such as fragmented lit and dark venues, different order book conventions, different benchmarking standards. Both CEOs push back on that premise, arguing that the structural similarities are greater than they appear.
Humphrey notes that while terminology differs between FX and equities – last look versus conditional order types, ECNs versus dark pools – the limit order book is fundamentally the same across both. Bassi goes further, arguing that equities and futures are in some respects cleaner to work with: once you have a reliable source of market data, the engineering and standardisation challenge is more tractable than in FX, where even basic questions about volume are contested.
A year-long industry pilot
BMLL and Tradefeedr are inviting market participants to join a year-long pilot to co-develop and validate the new multi-asset analytics capability. Participants will work with both firms to define metrics, stress-test data quality, shape what the firms are calling “AI-ready context layers,” and provide feedback on benchmarks and reporting, all delivered through Tradefeedr’s existing network and legal framework.
Bassi is candid about why the timeline is so long. “Banks and buy-side clients have so much on their plates that they don’t have time to figure out on their own how to get the data in, how to clean it, how to get through governance, and how to start using the tools,” he says.
The pilot mirrors the approach Tradefeedr took when it launched in FX – a public, collaborative process designed to build adoption in waves as successive cohorts of participants contribute data and iterate on outputs.
Humphrey emphasises that the existing infrastructure lowers the barrier to entry. Tradefeedr already has standardised data-sharing agreements with its client network, meaning the legal and governance framework required for participants to share private trading data is already in place. “That framework is the heavy lift – years of Tradefeedr’s time and effort going around one by one to these clients, convincing the buy side to convince the sell side to release their data,” he says. “They’ve done an exceptional job of that.”Challenging the incumbents
The partnership positions Tradefeedr and BMLL squarely against established multi-asset TCA providers – including Bloomberg, Virtu and the analytics arms of the major banks – that already cover equities and futures. The pitch is that an API-native, network-driven model can deliver more flexible, more accessible analytics than the vertically integrated incumbents, particularly as firms begin integrating LLMs into their execution workflows.
Bassi does not shy away from that framing. “The industry is full of individual vendors, different commercial models, a cottage industry of point solutions, or incumbents that just keep adding on features and charging a bit more without genuinely opening things up,” he says. “They’re in defensive mode, not in aggressive expansion mode.”
Whether that confidence is justified will depend on execution. The partnership is still at the pilot stage, the AI-ready analytics layer is being co-developed with the industry rather than delivered as a finished product, and the competitive response from incumbents with deeper coverage and larger installed bases remains to be seen. But the strategic logic is clear: as AI-driven workflows create new demand for clean, accessible, API-delivered trading data, the firms that control that connective layer between raw data and the front office will hold significant leverage. BMLL and Tradefeedr are betting they can get there first.
Subscribe to our newsletter
