QuantumStreet AI
QuantumStreet AI sells AI driven investment research, signals and index construction to institutions. It operates two models: portfolio as a service, delivering AI enhanced indices, thematic baskets and portfolios that institutional clients license and build products on, and software as a service, where clients use components of its Watson integrated platform including forecasts, signals, news analysis and knowledge graphs. Named index products include AIPEX, AIGO, AISRT and DBIQ. It is an IBM partner and the institutional division of EquBot, a registered investment adviser. Buyers are investment banks, asset managers, wealth managers, hedge funds and allocators including pensions, endowments, foundations and insurers.
Capability Axes
Capability grades
15 of 15 axes rated · 7 graded A or B
The removal test leaves nothing. The product is a set of index and portfolio methodologies generated by deep learning models that integrate fundamental, technical, macroeconomic and news data, plus a signal and research platform delivering forecasts, news analysis and knowledge graphs. There is no underlying rules engine, workflow product or data platform that would survive the models being taken out, because the models are the methodology.
The company was building and deploying machine learning models for institutional portfolios from 2016, which places it among the longest running model first vendors in this index rather than a platform that added a model layer later.
The company states that its platform automates the design, testing and ongoing analysis of investment strategies, which places model output at the point where a methodology is created rather than merely executed. The mitigating property is real: the resulting strategies are described as transparent and rules based, so once an index is published its behaviour is deterministic and observable, which is a stronger constraint than most agentic products offer.
What is missing is the step in between. No approval gate, investment committee review, methodology sign off or challenge process is described for the point at which a machine designed strategy becomes a live index that institutional capital tracks, and no account is given of who is accountable for a methodology no person authored.
Two mechanisms are genuine product design rather than assertion. Explainability is named as a core property throughout, with signals described as transparent and traceable to fundamentals, technicals, macroeconomic factors and news rather than delivered as an opaque score.
And the platform lets an institution build, backtest and launch a tailored index, so a prospective methodology can be tested against history before capital tracks it, which is the same test before deploy capability that earned Taktile its grade. Against that sits the gap that matters most for an investment product.
Performance is asserted repeatedly and never made checkable: no benchmark named, no period beyond a quarter, no methodology, and no statement of whether reported outperformance is live or simulated. Backtested index performance is the most abused figure in this market and the vendor publishes nothing that would let a buyer distinguish it.
Roughly a decade of deployment, an IBM partnership documented in IBM's own partner directory, institutional client events hosted at IBM sites in London and Munich, and four named index products, AIPEX, AIGO, AISRT and DBIQ. The scale claim is assets linked to its indices rather than a customer count, which is the right unit for an index provider. Two cautions belong on the record.
The figure moves materially across sources and dates, appearing as 5 billion, 7 billion, 7.5 billion and 8 billion dollars within roughly a year, so the Sardine precedent applies and any citation should be anchored to a dated source. And no institutional client is named anywhere, so the assets linked figure cannot be attributed to any identifiable buyer. Performance claims of broad outperformance across strategies are carried in paid press releases.
The stewardship question here is unusual and it is unaddressed. QuantumStreet is the institutional division of a registered investment adviser, which means the affiliated group manages money using models of the same family it licenses to institutional clients, and those clients invest in the same markets.
Nothing published describes separation between the adviser's own investment activity and the signals, indices or research delivered to licensees, how sequencing is handled where an index rebalance is known in advance, or whether client portfolio data observed on the platform is walled off from the affiliate. This is the dual position shape already recorded for TRM Labs and Rogo, appearing here inside a single corporate group rather than across a customer base.
The privacy chain is structurally shorter than most of this index because the inputs are market data, fundamentals, macroeconomic series and news rather than personal information about individuals, the same mitigating property recorded for Nammu21.
The exposure that does exist is different in kind: where an institutional client uses the platform for portfolio construction and monitoring, its holdings and intended positioning pass through the vendor, and that is material non public information whose leakage moves markets rather than harming a consumer. Nothing published addresses handling, segregation or retention of client portfolio data, which is the Rogo shape of the question.
No attestation, certification, trust centre or dedicated security page was located, and no service organisation control report or international information security standard certificate is announced or offered on request. The IBM partnership means the underlying infrastructure carries its own assurance programmes, but that is the platform provider's posture rather than the vendor's, and an institutional buyer's diligence process asks about the vendor. For a firm handling client portfolio positioning, the absence is the item that surfaces first in any operational due diligence questionnaire.
This vendor carries a real registration rather than a stated posture: QuantumStreet is described as the institutional division of EquBot, a registered investment adviser, which is a formal admission process passed with a securities regulator and the strongest licensure fact in this index after Norm Ai's affiliated law firm. Under the index convention that is what lifts the axis above the technology supplier default.
Two applicable regimes are unaddressed and both bear directly on the business. Published claims of outperformance across strategies engage the advertising and performance presentation rules that apply to registered advisers, and nothing distinguishes live from simulated results or names a benchmark or period.
Separately, an index administrator supplying benchmarks to European institutions falls under the European benchmark regulation and its authorisation requirement, which sits awkwardly beside a global leadership claim and a partner directory listing that says United States only.
No consumer subjects and no protected class analysis applies, so the axis adapts to the product as it did for Hadrius and Federato. What replaces it is methodology governance, and that is where the exposure sits. Thematic indices require a model to decide which companies belong to a theme such as artificial intelligence infrastructure or advanced manufacturing, and that classification directs institutional capital toward some issuers and away from others at scale.
Sustainability driven portfolios encode contested value judgements in the same way. Nothing published describes who governs the classification taxonomy, how a disputed inclusion or exclusion is reviewed, or what recourse an issuer has when a model determines it does not belong in a theme. Explainable output describes how the model reasoned, which under the standing rule set with Sardine is not the same as evidence the outcome is defensible.
No guarantee or indemnity on model output was located and none would be expected for an investment product, where losses are the client's by definition. What earns the grade is structural accountability of the kind recorded for Norm Ai: the affiliated entity is a registered investment adviser, which carries fiduciary obligations, examination exposure and a regulator with enforcement powers over how models are used and how performance is presented.
That is a real constraint no unregistered software vendor in this index operates under. The limit is that the fiduciary duty attaches to the adviser relationship rather than to an index licence, so an institution licensing a methodology and building a product on it sits outside that protection, and nothing published describes correction, restatement or notification if an index methodology is found to have been miscalculated.
The infrastructure and model platform dependency is named openly and corroborated by the provider itself, which is uncommon in this index. The company describes a Watson integrated platform, identifies itself as an IBM partner throughout its material, and appears in IBM's own partner directory, so a buyer can establish whose technology sits underneath rather than inferring it. That is the disclosure this axis exists to elicit.
What remains undisclosed is the data side, which for an investment signal product is half the chain: no market data vendor, news provider, fundamental data source or alternative data supplier is named anywhere, despite the platform's core claim being the integration of exactly those inputs.
One infrastructure relationship is named and substantiated, the IBM partnership, with the platform described as Watson integrated and the company listed in IBM's own partner directory with delivery stated as software as a service. Beyond that, integration is thin. No order management system, portfolio management platform, custodian, index calculation agent or market data vendor is named, and no developer documentation or application programming interface reference was located.
For the licensing model that matters less, since an index reaches a client through published methodology and calculation rather than through integration, but an institution taking the signals and research route would need to know how they arrive in its own environment and nothing published says.
Delivery is stated as software as a service in the IBM partner listing and a third party company record indicates United States server location, but neither is a vendor residency commitment and no region selection, private deployment or hosting detail appears in the company's own material.
That is a live question given the same listing states support for the United States only while the company markets to global banks and allocators, since a European or Asian institution consuming signals would need to know where its portfolio data comes to rest.
No pricing, licensing fee basis or packaging is published for either the index licensing model or the platform subscription, and every route in is a demo request. The basis of charge is the substantive question for an index provider, since fees are conventionally struck on assets linked to the index and that structure determines the total cost of a product built on it. Nothing indicates whether the two offerings are priced separately or bundled.
The named buyer set is wide and consistent across sources: investment banks, asset managers, wealth managers and hedge funds on the manager side, and pensions, endowments, foundations and insurers on the allocator side. Both consumption models are offered, so an institution can license a finished index or take the underlying signals and build its own, which widens the set further.
One inconsistency belongs on the record: the company describes itself as a global leader with strategies deployed globally, while its own listing in the IBM partner directory states the regions and countries supported as the United States only. Those two statements cannot both be complete.
Compared With
Most editorial comparisons pair two vendors the index assesses as direct competitors for the same buyer. Some pair vendors that are adjacent rather than rival, where the useful question is where one ends and the other begins. Each carries a verdict, the buyer conditions that favor each vendor, and a graded side by side.
Alternatives to QuantumStreet AI
The closest documented capability profiles to QuantumStreet AI in the same categories, ordered by similarity across the same fifteen axes the index grades every vendor on. Closest documented profile, not a claim that either product does the same job. No vendor pays for placement.
Stronger documented coverage on Model Risk Management and Transparency
Documents GLBA and Data Privacy Posture and Autonomy and Oversight Model, among others where QuantumStreet AI does not
Documents AI Safety and Data Stewardship where QuantumStreet AI does not
Documents Autonomy and Oversight Model where QuantumStreet AI does not
Documents Autonomy and Oversight Model and Core Systems and Integration Depth where QuantumStreet AI does not
Documents GLBA and Data Privacy Posture and Autonomy and Oversight Model, among others where QuantumStreet AI does not
Similarity is computed axis by axis from published grades, not from a composite score. The index does not aggregate grades into a total. See the fifteen axes and the methodology.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
No pricing data has been verified for this vendor. Pricing information will be published here once confirmed through vendor disclosure or third-party estimation.