Hummingbird acquires Quantimon to connect machine learning with AML transaction monitoring workflows
Oct. 29, 2026 • Article • Companies & Deals
Hummingbird announced its acquisition of Quantimon on October 28, 2026, as it works to connect machine-learning risk detection with compliance investigations. Financial terms were not disclosed. Quantimon’s team has joined Hummingbird to build closed-loop transaction monitoring capabilities, while co-founder and CEO Ben Duranske will advise during the transition. The companies had been commercial partners for approximately a year.
The proposed feedback loop would use Hummingbird’s structured records of alert resolutions to retrain Quantimon’s detection models. Quantimon claims its algorithms can reduce false-positive alerts by up to 67%; the research provided does not establish how that figure was measured. Its federated learning network is designed to help institutions identify emerging financial crime patterns without sharing underlying customer or transaction data. The deal is Hummingbird’s second acquisition, following its purchase of data automation and integration startup LogicLoop in September 2024.
Hummingbird announced on October 28, 2026, that it had acquired Quantimon to bring machine-learning risk detection into its anti-money laundering (AML) transaction monitoring workflows. Financial terms were not disclosed. The aim is to connect what compliance teams learn while investigating alerts with the models that identify potentially risky activity.
Quantimon’s team has joined Hummingbird to build closed-loop transaction monitoring capabilities. Quantimon co-founder and CEO Ben Duranske will remain an adviser during the transition. The companies had operated as commercial partners for approximately a year before the acquisition.
Turning investigations into model feedback
The planned approach depends on information created after a transaction monitoring alert is raised. Hummingbird’s platform records alert resolutions made by compliance analysts and AI agents in structured formats. Those records can provide labeled training data for retraining Quantimon’s machine-learning detection models, creating a feedback loop between investigations and future monitoring.
That connection addresses a practical obstacle in AML monitoring. Traditional systems have largely relied on static, rules-based logic because the outcomes of earlier investigations were often buried in unstructured compliance case notes. Converting those notes into training datasets could be prohibitively expensive. Recording resolutions in a structured form makes the results of investigations more usable for model development.
A closed loop in this context means that monitoring does not stop at generating an alert and recording a decision. The decision becomes input for subsequent training of the detection models. Hummingbird and Quantimon intend to build that capability; the acquisition announcement does not establish that it has already been completed or deployed across Hummingbird’s client base.
Quantimon’s detection and data-sharing approach
Quantimon claims its AML risk detection algorithms achieve up to a 67% reduction in false-positive alerts. False positives are alerts that prompt review but do not reflect the risk the system was looking for. Reducing them could lessen the volume of cases requiring attention, although the supplied information does not specify how Quantimon measured its claim or whether that result would carry over to every institution.
Quantimon also incorporates a federated learning network designed to let institutions identify emerging financial crime patterns without sharing the underlying customer or transaction data. Federated learning generally refers to an approach in which participants can contribute to model learning without pooling their raw records. In Quantimon’s case, the stated purpose is to draw on patterns seen across institutions while keeping the underlying data with each institution.
That network is separate from, but relevant to, the proposed feedback loop. One concerns how institutions can learn from patterns without exchanging underlying data; the other concerns how resolutions of individual monitoring alerts can inform later model training. The supplied information does not detail how the two capabilities will be combined in Hummingbird’s product.
The acquisition in context
Hummingbird provides compliance orchestration tools to enterprise and fintech clients, including Raymond James, Crypto.com, DraftKings and Etsy. Its existing role in compliance workflows is central to the deal: the structured alert outcomes recorded on its platform are the potential training inputs for Quantimon’s models.
Duranske previously led compliance and risk functions at Apple, PayPal and Xoom. His advisory role gives the companies continuity during the transition, while Quantimon’s team takes on the work of building the new monitoring capabilities within Hummingbird.
The purchase is Hummingbird’s second acquisition to date. In September 2024, it acquired LogicLoop, a data automation and integration startup. The Quantimon deal adds a detection-model component to Hummingbird’s compliance technology, following a year-long commercial relationship between the two companies.
The next step described in the acquisition announcement is development of closed-loop monitoring, not a stated rollout date or a reported customer deployment. Its potential value rests on whether structured investigation outcomes can be used to keep detection models current while preserving the controls around sensitive financial data. For now, the confirmed development is the acquisition and the transfer of Quantimon’s team to that work.
Hummingbird acquires Quantimon to connect machine learning with AML transaction monitoring workflows
Hummingbird announced its acquisition of Quantimon on October 28, 2026, as it works to connect machine-learning risk detection with compliance investigations. Financial terms were not disclosed. Quantimon’s team has joined Hummingbird to build closed-loop transaction monitoring capabilities, while co-founder and CEO Ben Duranske will advise during the transition. The companies had been commercial partners for approximately a year. The proposed feedback loop would use Hummingbird’s structured records of alert resolutions to retrain Quantimon’s detection models. Quantimon claims its algorithms can reduce false-positive alerts by up to 67%; the research provided does not establish how that figure was measured. Its federated learning network is designed to help institutions identify emerging financial crime patterns without sharing underlying customer or transaction data. The deal is Hummingbird’s second acquisition, following its purchase of data automation and integration startup LogicLoop in September 2024.
Guidewire Software has agreed to acquire hyperexponential for $1.7 billion in cash, with up to $400 million more payable if specified post-closing recurring revenue targets are met. The agreement was announced on October 27, 2026. Subject to customary regulatory and antitrust conditions, the transaction is expected to close in the second quarter of Guidewire’s fiscal 2027. The deal would bring hyperexponential’s pricing and underwriting software, including hx Renew and its hyperoperator agentic workbench, into Guidewire’s business. The platform supports decisions across more than $75 billion in annual commercial property and casualty gross written premium. At Guidewire’s Investor Day, CFO Jeff Cooper projected the acquisition would add at least 650 basis points to fiscal 2027 annual recurring revenue growth inorganically, followed by at least 150 basis points of organic ARR expansion annually from fiscal 2028. Guidewire plans to update its fiscal 2027 guidance in December 2026.
Whop announced a $200 million Series D on October 21, 2026, valuing the digital commerce platform at $5.1 billion. Sequoia Capital led the round, with Shopify Ventures participating. The valuation is about 3.2 times the $1.6 billion figure set during Tether’s strategic funding round in February 2026. Former Sequoia leader Doug Leone joined Whop’s board following the announcement. Whop said its annualized run-rate revenue exceeded $400 million as of September 2026, compared with $60 million at the beginning of 2025. It reported roughly 25,000 active business operators and an average blended transaction fee of approximately 6%. The company plans to expand its international team and establish regional offices in Europe and Colombia. It also reported more than 350,000 daily API calls across over 14,700 monthly active users of its newly launched Whop CLI and Model Context Protocol connectors, pointing to growing use of its developer tools.
London Stock Exchange Group has agreed to acquire credit intelligence provider Octus for $3.535 billion in cash. LSEG plans to fund the purchase entirely with external debt and expects the transaction to close in the first half of 2027, subject to customary regulatory approvals. Octus will join LSEG’s Data & Analytics division, with founder and CEO Kent Collier continuing to lead the business. The acquisition would add Octus’s private covenant and distressed credit data to LSEG’s debt-market offerings. LSEG plans to connect that information with products including Workspace, LPC and Yield Book, alongside Tradeweb. Octus is projected to generate more than $300 million in 2026 revenue at an adjusted EBITDA margin of approximately 40%. LSEG expects the deal to be neutral to adjusted earnings per share in its first full year after closing and accretive thereafter. It forecasts leverage reaching the upper end of its target range by the end of 2027.
CRB Group, the parent of Cross River Bank, agreed on October 20, 2026, to acquire Accrue, a fintech focused on merchant-branded digital wallets and customer loyalty. Accrue founder and CEO Michael Hershfield and his team are expected to join Cross River to lead its embedded wallet and loyalty offerings when the deal closes. Financial terms and a completion date have not been disclosed, and the transaction remains subject to customary closing conditions. The companies already had a banking relationship: Cross River held deposited funds and issued Visa debit cards for Accrue programs. The acquisition would bring Accrue’s wallet capabilities into that existing arrangement. Closed-loop wallets can keep payments and refunds within a merchant’s branded balance, offering an alternative to routing every transaction through card networks. The deal also illustrates why embedded finance providers are looking beyond payment-processing margins toward wallet software and stored-value services, though the companies have not disclosed the acquisition’s expected financial impact.
Hyperlayer has raised £19.5 million in a Series A2 round led by existing investor Flintlock Capital, valuing the company at £195 million post-money. Mouro Capital and Susquehanna Private Equity Investments also participated. The funding brings Hyperlayer’s capital raised over the trailing 12 months to approximately £50 million, following a £30 million financing in late 2025. The company develops Layer Core, a programmable overlay designed to let banks manage product logic, pricing and automated compliance above their existing core systems, which continue to record ledger transactions. Hyperlayer says it has secured deployment mandates with unnamed Tier 1 banks in Europe and North America. The round supports an international rollout following its US market launch in September 2026. Alongside the financing, Hyperlayer named Ben Helps chief operating officer and Louise Chan chief commercial officer. Flintlock Capital co-founder and general partner Joe Shamess joined its board.