Whale Rounds Its Series C to $100 Million
Singapore-based Whale raised a $40 million Series C3 extension on July 16, pushing its total Series C to $100 million and signaling that enterprise AI for physical operations has real institutional backing. The extension was led by CMB International and SMBC's Asia Rising Fund, with a strategic roster that includes Hyundai Motor Group, Bosch Ventures, Singtel Innov8, Krungsri Finnovate, and others. We pay attention when industrial and banking investors write these checks, because that syndicate looks different from the usual growth-fund crowd. Their presence suggests the buyers are their own portfolio companies and clients, which is a more grounded demand signal than another consumer AI round chasing usage metrics.
Whale's pitch is an AI operating system, or AIOS, for enterprise operations, built on what it calls a Business World Model. Founder and chief executive Jerry Ye frames the platform as the connective layer between digital workflows and the physical world. The company already claims more than 1,600 enterprises across 45-plus countries and over 600,000 edge AI nodes in production, spanning retail, automotive, food and beverage, manufacturing, and financial services. Those are large numbers for a company most Western CIOs have never heard of, and they hint at how much enterprise AI adoption is happening outside the US software narrative.
A Model Built to Read Cameras and Audio
The technical premise is the interesting part. Whale describes its Business World Model as an AI designed to interpret signals from cameras, sensors, and audio the way large language models process text. That framing targets a gap in most enterprise AI programs, which are fluent in documents and databases and blind to what actually happens in a store, a showroom, or a factory. Mayoran Rajendra, head of SMBC's AI Transformation Office, said Whale's ability to unlock and structure data from physical environments was particularly compelling. For operations leaders drowning in unstructured video and sensor feeds, a model that turns that noise into structured signal is a concrete capability.
Two products carry the pitch. SpaceSight converts cameras and IoT sensors across physical sites into real-time intelligence on foot traffic, dwell time, engagement, and compliance. Echo analyzes frontline sales conversations to surface what top performers do and turn it into coaching. We would note that both attack a long-standing blind spot: the frontline, where most retail and manufacturing value is created and where digital transformation has historically stopped at the back-office door. If the accuracy holds, this is the kind of operational visibility that spreadsheets and quarterly audits never delivered, and it arrives continuously rather than after the quarter closes.
The Frontline Data Problem Is Finally Addressable
For years, the frontline has been the least instrumented part of the enterprise. Head office runs on dashboards while stores and plants run on manager intuition and periodic audits. Whale's bet is that computer vision and audio intelligence have matured enough to close that gap without deploying an army of analysts. Zheng Xiang of Charisma Partners described the platform as a perception-cognition-execution loop that reshapes the operational foundation for enterprises. Stripped of the venture gloss, that means sensing what happens, interpreting it, and acting on it in something close to real time. That loop is exactly what frontline operations have lacked, and it is where the durable value would sit.
We would still press on the hard questions any physical-AI deployment raises. Camera-based intelligence carries privacy and labor-relations exposure that a back-office agent does not, and 600,000 edge nodes is a large attack and governance surface. Buyers in regulated markets will need clear answers on data residency, consent, and worker monitoring before scaling. A technology being capable is one thing; the same technology being deployable inside your compliance regime is another. Leaders drawn to the operational upside should size the governance work honestly at the start, because retrofitting privacy controls onto a live camera network is far harder than designing them in.
Governance Is a Named Module, Which Is Telling
Whale's platform includes six components, and the one we find most revealing is Novus, described as AI infrastructure and governance, sitting alongside workflow automation, content, and knowledge modules. Governance shows up as a first-class part of the product, which tracks the broader enterprise shift we keep flagging: control planes are becoming the point of the sale. When a vendor leads with a governance layer, it is responding to buyers who learned in 2025 that ungoverned AI turns into shadow IT. For a CIO evaluating any operational AI platform, the presence and depth of that layer is now a primary selection criterion rather than a checkbox at the end of procurement.
The strategic investors reinforce the operational, regulated posture. SMBC and CMB International bring banking scrutiny, Hyundai and Bosch bring industrial and automotive deployment reality, and Singtel and Krungsri point to telecom and Southeast Asian financial services. These are not tourists in enterprise AI; they run large physical operations and strict compliance regimes. When investors like these fund a platform, they are also implicitly validating its governance and security claims against their own standards. We read the syndicate as a signal that Whale's control and compliance story survived diligence from buyers who cannot afford to get it wrong.
The Geography of Enterprise AI Is Widening
Whale plans to use the capital to expand from its current North American and Asia-Pacific base into the Middle East, Europe, and deeper across Japan, Indonesia, Malaysia, and Thailand. That map matters for Western leaders because it shows serious enterprise AI capability accumulating outside the familiar US vendor set. A platform with 1,600 enterprise customers and heavyweight Asian industrial backing is a credible global competitor, and its strength in physical operations is precisely where American software has been weakest. We would not assume the incumbents on your shortlist are the only viable options for frontline and edge use cases in 2026.
This widening also has a sovereignty dimension. Enterprises increasingly want operational AI they can run under their own jurisdiction's rules, and a diversified vendor landscape gives them leverage on data residency and price. For a CIO, the practical value is optionality. A market with strong providers across regions is harder for any single vendor to lock up, which strengthens your negotiating position and lowers concentration risk. Whale's raise adds a serious name to the frontline-operations category, and leaders building a 2026 shortlist for physical-AI use cases should make sure it appears on theirs.
What This Means for Operations Leaders
The signal from this round is that AI is moving off the screen and onto the floor, and the money is following it into stores, plants, and showrooms. If your transformation program has treated frontline operations as too messy to instrument, platforms like Whale are arguing that the excuse has expired. The practical first step is to pick one measurable frontline problem, out-of-stocks, compliance lapses, or inconsistent selling, and test whether continuous vision and audio intelligence beats your current periodic checks. Treat it as an operations pilot with a clear metric, staffed by the people who own the outcome, not as a technology showcase for the innovation team.
The paired obligation is governance, which has to be designed in from the first camera. Physical-AI deployments touch employees and customers directly, so privacy, consent, and worker-monitoring policies belong in the pilot charter, not a later compliance review. We would insist on the same controls Whale itself markets through Novus: clear ownership, audit trails, and defined limits on what the system watches and records. The upside of instrumenting the frontline is real, and this round shows investors believe it. Capturing that upside without a labor or privacy backlash is the actual work, and it starts before the first node goes live.
