SAP's €1B Prior Labs Deal Rewrites Structured-Data AI Competition

SAP Buys Native AI for Structured Data

SAP announced it will acquire Prior Labs, a German frontier-AI lab, for approximately €1 billion (~$1.16 billion) over four years. The deal brings Prior Labs' TabPFN tabular foundation models — downloaded more than 3 million times as open-source tooling — directly into SAP Business Data Cloud. This is the clearest signal yet that large platform vendors are abandoning reliance on third-party model integrations in favor of native AI capability for enterprise structured data.

The competitive implication is immediate: Microsoft, Salesforce, Oracle, and Databricks all rely heavily on partnerships with OpenAI, Anthropic, or other model providers to power structured-data workflows. SAP is instead embedding model capability into the data platform itself, which reduces latency, simplifies governance, and tightens lock-in. For enterprise buyers evaluating ERP or data-platform deals, this means stronger native AI automation for structured analytics, forecasting, and tabular workflows — but also higher lock-in risk because AI features become inseparable from SAP's data stack and governance model.

Real-Time and Agentic Infrastructure Takes Capital

LiveKit raised a $100 million Series C on January 22 to scale its real-time voice and video infrastructure for voice-first and agentic applications. The company is positioning against Twilio, Agora, and Vonage, but the market is shifting from generic communications APIs to infrastructure designed for always-on AI agents and multimodal user experiences. Contact centers, sales, and support automation budgets are increasingly funding real-time AI experiences that require low-latency voice and video primitives, not just chat or batch workflows. Buyers should expect vendors to ask for separate infrastructure spend to support agentic UX, which adds complexity to procurement scopes.

DealHub raised a $100 million growth round the same day to accelerate its agentic quote-to-revenue platform, targeting Salesforce CPQ, Conga, and PandaDoc. The competitive edge is moving from static CPQ rules to AI-assisted quote generation and workflow automation. Procurement teams will face more pressure to evaluate whether their quote-to-cash stack can support AI-assisted selling and pricing without adding manual RevOps headcount. This is a proxy for a broader shift: vendors are rebuilding around agent workflows rather than extending classic seat-based SaaS.

Pricing Models Are Breaking

Vendors are abandoning seat-only SaaS in favor of consumption and outcome-based pricing. An analysis cited by SaaS Intelligence shows vendors increasingly combine subscriptions with usage charges and outcome-based pricing tied to completed AI tasks. This weakens traditional per-seat incumbency advantages for Microsoft, Salesforce, and other large suites, while rewarding vendors that can meter agent activity and AI throughput.

Budget owners should expect more variable spend, new governance controls, and harder contract negotiations because AI usage creates unpredictable run-rate costs. Zylo expanded its platform to manage both traditional SaaS subscriptions and AI consumption spending in one system, reflecting a new buyer requirement for consolidated visibility. CIOs and procurement teams will increasingly need unified controls over software licenses and AI tokens to avoid budget surprises and compliance blind spots. This extends SaaS management into FinOps-like territory and pressures Flexera, Torii, Vendr, and Apptio-adjacent tooling to support AI usage controls.

Other Material Deals

Claroty raised a $150 million Series F on January 22 to support global expansion and platform development in cyber-physical systems protection. The company competes with Armis, Nozomi Networks, and OT-security offerings from major security suites. For buyers in manufacturing, utilities, and healthcare, this supports larger-scale OT/IoT security deployments, but it also raises expectations that platform vendors can unify visibility, risk scoring, and incident response across plant and enterprise environments.

Pennylane raised €175 million (~$205 million) on January 20 for accounting and spend management expansion, including AI product investment. The company is pushing against QuickBooks, NetSuite, Sage, and spend-management tools like Coupa, with differentiation centered on AI-assisted finance workflows and European market expansion. Finance leaders will see more integrated accounting-plus-spend platforms that can reduce tool sprawl, but also face vendor concentration risk if they consolidate core finance workflows into one SaaS layer.

Deel acquired Sastrify, bringing a $45.3 million-funded SaaS procurement and license management platform into Deel IT. This moves Deel beyond payroll and hardware lifecycle into SaaS procurement and spend optimization, competing more directly with Zylo, Torii, and Vendr. Finance and IT buyers may get a more unified vendor-management workflow, but should weigh the tradeoff between operational convenience and dependence on a broader workforce platform for SaaS governance.

What to Watch

Datadog acquired Adaptive ML to extend beyond observability into Reinforcement Learning Operations for enterprise AI. This puts pressure on New Relic, Dynatrace, and emerging AI ops vendors to offer comparable model-tracing and agent-control capabilities. Buyers are increasingly treating AI observability and model governance as part of infrastructure spend, not an optional add-on, which changes procurement scopes and security review requirements.

The clearest enterprise-buyer signal across all these deals is that vendors are rebuilding around agent workflows, consumption-based pricing, and AI controls rather than classic seat-based SaaS. Buyers should plan for more variable spend, tighter governance requirements, and harder contract negotiations as AI usage becomes the dominant cost driver in SaaS infrastructure.