The clearest signal in this quarter's enterprise software results is that the money is moving toward the layer that supervises AI agents, not only the layer that builds them. On 15 October, ServiceNow announced an AI Passport that verifies the scope of first-party and third-party AI agents in real time, alongside an expanded governance product it calls the Super Intelligence Control Tower.1 Its Jon Sigler framed the pitch plainly: "As AI agents proliferate across the enterprise, visibility alone isn't enough. Enterprises need confidence that every AI agent is operating within its authorized scope."1
Governance becomes a product
The ServiceNow announcement is a bet that agents will be deployed faster than companies can audit them. Its stated aim is to unify agent authority, governance and operational oversight so enterprises can scale agentic AI with control.1 In our entity data, the ServiceNow AI Platform is built on NVIDIA's AI-Q, Agent Toolkit and Nemotron, so the governance layer sits on top of the same GPU vendor's agent stack that its customers are adopting.2
Confluent sits on the data side of the same shift. Our data records its Data Streaming Platform as built on Streamhouse, and Lightning Tables as built on top of that platform.2 Our dossier holds no Confluent financial figures, so we can say what it built but not how the market has rewarded it. Qualys, whose ETM product is recorded as developed by Qualys, Inc., is the one security vendor in this group where the numbers are in hand.2
What the Qualys numbers show
Qualys reported third quarter 2026 revenue of $188 million, up 11% year over year, with adjusted EBITDA of $89.9 million (a 48% margin), EPS of $2.12 and free cash flow of $100 million (a 53.2% margin).3 It then raised full-year revenue guidance to $736.9 million to $738.9 million, from $732 million to $738 million, and raised full-year EPS guidance to $8.07 to $8.14, from $7.74 to $7.88.3 The company guided fourth quarter revenue to $191 million to $193 million, which it describes as 9% to 10% growth.3
Note the shape of the raise. The revenue range moved only at the bottom, and its top end rose by under $1 million. The larger move was in earnings per share. That is a company becoming more profitable on steady growth, not one riding a surge in AI demand. We would not read it as proof that AI agents are driving security spending.
Where the investors see an opening, and where they do not
The most useful practitioner view in the dossier comes from Sean Jacobsohn of Norwest, who spent years in HR software before investing in it.4 He is blunt about the limits of AI in the finance stack: "You don't want AI doing calculations because it is not good at math. There are certain workflows it can handle where it doesn't produce precise numbers. But when you need precision, accuracy and calculations, you can't rely on AI for that."4
That matters for anyone building agents for accounting or payroll. The opening he sees is not the core product of the incumbents: "I think it would be very hard to disrupt the core products of Workday, ADP, SAP, UKG and Dayforce. But it's easier to disrupt some of their secondary products, where the category isn't their core business."4 He also points to a sales mechanic: "There's actually one less layer of approval when they're buying their own software, so it is a little easier to replace it when they're the direct buyer."4 His test for founders is similarly practical: "Before I invest, I'll go on a lot of sales calls I set up with the CEO to see how good they are at selling."4
Read together, these describe a narrow path for AI-native challengers: attack non-core modules, sell to the CFO directly, and keep the language model away from the arithmetic.
Agents in the field, and the small-cap squeeze
Dextr AI, which builds agents for hotels, raised $6.7 million, and says its agents now handle more than 1 million interactions a month across hundreds of hotels.5 That usage figure is the company's own claim, and the source, Crunchbase News, has a measured reliability of 49% of 41 checked claims holding up. Treat the number as unverified.
At the small end, AI is showing up as a cost tool. TEN Holdings reported third quarter revenue up 17.9% to $640,000.6 Chief executive Virgilio Torres said the company cut its annual operating cost structure by about $2.5 million over 12 months, partly through "the integration of AI-driven tools to streamline engineering and accelerate product development."6 It also cut short-term related-party borrowings from $4.6 million to about $0.1 million, and total liabilities by roughly 74%, from $7.7 million to $2.0 million.6 The quarter included a $1.8 million write-off of previously capitalized service contracts and about $0.5 million of legal expenses, which the company calls non-recurring.6 Those two items are larger than the quarter's revenue, so the headline growth should be read with them in view.
Consolidation is the other small-cap story. Tiny Ltd. completed its acquisition of Oso Cloud, an authorization and access control software business, for about $1.9 million in cash.7 It adds roughly $5.3 million of annual recurring revenue, lifting Tiny's pro forma total about 7.6%, from $70.0 million to about $75.4 million.7 Oso served about 80 customers in security, fintech and developer software, all on recurring subscriptions, including several multi-year agreements.7 Authorization is exactly the problem ServiceNow's AI Passport addresses at enterprise scale, which suggests the category is being valued from both ends.
Policy and the one outlook that slipped
On 4 November, the day after the midterm vote, CIBC Capital Markets published a report recommending a 36-stock Democratic basket, alongside a 45-name Republican basket built before the election.8 The topic summary we work from says CIBC expects large cloud and software platforms to benefit in a regulation-driven market. The report itself is not in our dossier, so we cannot say which companies are in it or why.
Trimble, on the same date, kept its full-year 2026 revenue midpoint at $3.925 billion (more than 9% organic growth) and nudged its EPS midpoint up to $3.66.9 But it cut its annual recurring revenue growth outlook to about 10%, from 12% to 14%.9 For a software-led business, subscription growth is the number the market watches most closely, so that is the real headwind.
What to watch
- Whether Qualys's raised EPS guidance holds when it reports the fourth quarter against its $191 million to $193 million revenue range.3
- Whether TEN Holdings' operating costs fall in the next quarter once the $1.8 million write-off and legal costs drop out.6
- Whether AI-native finance products stay on the workflows that need no precise arithmetic, as Jacobsohn advises.4
- Whether the CIBC baskets, once readable, actually name the cloud and software platforms the topic summary expects.8
On sourcing: the press-release documents here come from a feed where 54% of 6,692 checked claims held up, and the Crunchbase News pieces from one where 49% of 41 did. Company statements are the company's own account, and we have flagged where nothing independent confirms them.

