BestEx Research has launched Pulse AI, a conversational interface that allows institutional trading teams to access the company’s pre-trade analytics through ChatGPT and Claude, according to the firm’s launch announcement. The currently available product covers pre-trade analysis for global futures, while equities analytics and broker-neutral post-trade transaction cost analysis remain scheduled for later releases.
The launch gives traders a natural-language route into BestEx Research’s existing liquidity data and transaction-cost models. Instead of asking an analyst to extract information and construct a report, an authorized user can ask when a futures contract is normally most liquid, how its spread changes through the trading day, what a large order may cost at different execution speeds, or how scheduled economic data tends to affect order-book depth.
Pulse AI should therefore be understood as a conversational access layer over proprietary analytical models and processed market data. The language model helps interpret a question and present the results, but it does not independently create the underlying market-impact model. The quality of an answer will still depend on BestEx Research’s data, assumptions, instrument coverage, and analytical methodology.
Only the Futures Pre-Trade Component Is Live
The distinction between available functionality and the product roadmap is important. BestEx Research’s Pulse AI product page says pre-trade analytics are live for global futures. Equities pre-trade support is expected in October 2026, while broker-neutral post-trade analysis is scheduled for the first quarter of 2027 and is being offered through a beta program.
For futures traders, the live system can return estimated market impact, historical liquidity statistics, spreads, depth, and realized or estimated microstructure data. Those outputs could help a trader compare participation rates, identify more favorable trading windows, or determine whether a position should be executed quickly at a higher expected cost or worked more gradually with greater exposure to market movement.
BestEx Research says its broader Pulse cost model produced estimates within one tick of realized parent-order costs across a company-tested dataset of more than 72,000 orders and over 65 base symbols. That is a potentially useful indicator, but it is vendor-published testing rather than an independent audit of Pulse AI. The company has not yet disclosed separate accuracy statistics for answers produced through the conversational interface.
The product builds on BestEx Research’s expansion beyond execution algorithms. The company previously raised $10 million to expand into additional asset classes and regions, while its technology for brokers has been developed to support firms building their own branded algorithmic execution offerings. Pulse AI moves the interaction with those analytical capabilities into software that many institutional users already have open.
Broker-Neutral TCA Is the More Ambitious Part of the Roadmap
The planned post-trade component could address a genuine weakness in institutional execution analysis. A buy-side desk using several brokers often receives reports built from different datasets, benchmarks, and methodologies. Because each broker sees only the orders routed through its own systems, the resulting reports do not necessarily allow direct comparisons across the entire trading program.
BestEx Research intends to apply a common methodology to activity across brokers, strategies, instruments, and venues. In principle, that could help a trading desk identify which algorithms performed well under particular liquidity conditions and whether execution choices should change. A conversational interface could then let users investigate performance immediately instead of waiting for a static monthly or quarterly report.
Broker-neutral analysis is not method-neutral, however. Meaningful comparisons require consistent timestamps, parent-and-child order mappings, fee treatment, currency conversion, venue data, and benchmark definitions. Decisions about whether to measure arrival-price slippage, implementation shortfall, volume-weighted performance, or another benchmark can materially affect the outcome.
This is why partnerships such as the effort by BMLL and Tradefeedr to construct AI-ready cross-asset trading analytics focus heavily on normalized, high-quality data. A convenient conversational interface cannot compensate for incomplete execution records or inconsistent comparison methods.
The “World’s First” Claim Needs Qualification
BestEx Research describes Pulse AI as the world’s first AI-native interface for institutional trading analytics. That claim cannot be independently established without a narrow definition of the product category. Other providers have already introduced artificial intelligence into institutional execution analysis, including Broadridge’s AI-based algorithm insights service.
Tradeweb also launched TARA in June 2026, giving institutional credit traders a conversational interface for questions concerning liquidity, market activity, and execution performance. The arrival of Tradeweb’s conversational analytics for credit markets predates the Pulse AI announcement and demonstrates that natural-language institutional trading analysis is not unique to one provider.
Pulse AI may still be differentiated by its combination of futures-focused pre-trade cost modelling, planned broker-neutral analysis, and direct availability through general-purpose assistants such as ChatGPT and Claude. That is a more defensible description than an unrestricted claim to be the first AI-powered institutional analytics platform.
Connecting Trading Data to General AI Assistants Raises Governance Questions
BestEx Research says an active Pulse account is required and that the connection is currently supported through ChatGPT and Claude. Although the announcement refers more broadly to other assistants, the company’s current product FAQ specifically identifies those two platforms. Enterprise and team deployments normally require an administrator to add the custom connector for the organization.
The company says a user’s existing password is used once to issue an access key and is not stored by the connector. Institutional customers will nevertheless need to examine how prompts, outputs, identifiers, and trading information are handled across the full workflow. Permissions should prevent one user from accessing another desk’s orders, while logs must show which data was retrieved and how an analytical answer was produced.
Firms will also need controls for inaccurate or unsupported interpretations. A language model can make a complex report easier to interrogate, but a fluent answer is not automatically a correct one. Material execution decisions may still require the user to inspect the underlying data, assumptions, visualizations, and confidence range.
Pulse AI is presented as an analytics product rather than an order-entry system. That separates it from initiatives such as TradeStation’s connection between trading accounts and AI platforms, where the potential scope extends closer to account actions and execution. The narrower analytical boundary may reduce operational risk, provided the connector cannot initiate trades or modify orders.
“It’s a fundamentally different relationship with the data than a quarterly report can produce,” said BestEx Research founder and CEO Hitesh Mittal.
Adoption Will Depend on Measurable Trading Outcomes
The immediate value of Pulse AI is speed of access. If traders and quantitative analysts can retrieve reliable answers without repeatedly assembling data extracts and reports, the system could shorten the path between observing market conditions and selecting an execution approach. It may also make sophisticated analytics accessible to users who do not write database queries or analytical code.
The stronger institutional case will require evidence that the interface saves analyst time without weakening oversight and that its recommendations contribute to lower execution costs. Customers will also want consistent outputs, documented methodologies, clear entitlements, and an auditable link between each answer and its supporting data.
For now, Pulse AI is a live conversational front end for global futures pre-trade analytics with a broader suite on the roadmap. Equities coverage and broker-neutral post-trade TCA could turn it into a more complete execution-analysis environment, but those capabilities should be evaluated when they become operational rather than treated as features of the present release.