SAP CTO Philipp Herzig put the whole month into one sentence: "Enterprise AI doesn't stall because the models aren't good enough — it stalls because the data isn't ready for AI agents." That quote explains both July acquisitions. Dremio (closed July 6) eliminates the ETL problem — SAP and non-SAP data, mixed in real time, no movement required. Prior Labs (closed July 17, €1B+ committed over four years) brings Tabular Foundation Models that predict payment delays, supplier risk, and churn from structured ERP data — something LLMs have never done well. Together: a data lakehouse feeding purpose-built AI models feeding Joule agents. The stack is complete.

Per-user pricing is officially dead. SAP CEO Christian Klein told Bloomberg this month: "It would be foolish to still charge subscription base, because AI is so powerful that it will automate a lot of tasks." Use-based pricing is now the default on all cloud renewals. If your clients haven't modeled what AI agent consumption actually costs at scale, they are about to find out at renewal time. Get that conversation in front of procurement now. (ERP Today)

Security — don't skip this. SAP Patch Day on July 14 dropped 16 notes including 4 critical. A serious vulnerability in SAP Approuter (BTP environments) exposes login callbacks via X-Forwarded-Host misconfigurations. Additional critical entries span ABAP, Java, Fiori, and SAProuter. If your clients are live on BTP workloads and haven't patched, flag it today. (Pathlock)

SAP issued €3.5 billion in new debt to fund its acquisition run, with the majority earmarked for scaling Prior Labs over four years. Bad data infrastructure wastes an estimated 08 billion in enterprise AI spend annually according to a Hitachi Vantara analysis cited by CIO Dive. SAP is betting it can capture a large slice of that waste-reduction market by owning the data layer end-to-end.

The detail most people missed in Klein's pricing announcement: SAP is also building "forward deployed engineering" teams — hundreds of employees who will co-develop AI workflows directly inside customer environments. This is SAP doing what SIs do. It's the carrot (€100M partner fund, 200+ agents) and the stick (internal consulting capability) at the same time. SI partners need a clear answer to the question customers will start asking: why do I need you when SAP is offering this directly?

Meanwhile, S/4HANA 2025 is the current on-premise release with maintenance through December 2030 — and the three-way fight over ECC migration strategy (lift-and-shift vs. brownfield vs. greenfield) is intensifying as the deadline pressure builds. Consultants who can make a clear, numbers-backed case for one path over another are in demand.

The partner ecosystem is being squeezed from two directions simultaneously. From above: SAP's forward-deployed engineering teams moving into implementation work. From below: AI tools like KTern.AI, which just launched on AWS Bedrock, specifically targeting the discovery, impact analysis, and test case generation phases of SAP transformations — historically the most billable consulting hours. (ERP Today)

SAP renamed the partner competition from "SAP BTP" to SAP Business AI Platform as of June 30 — a signal of where the platform identity is heading. The three-layer stack (context → build with Joule Studio 2.0 → govern with AI Agent Hub) is now the official partner build framework. If you're pitching BTP projects, your proposal deck needs to reflect this architecture.

The €100M SAP partner fund is still active. 50+ Joule Assistants and 200+ specialized agents are in the catalogue and available to build on. The opportunity is real — but the window for building differentiated IP before it becomes commoditized is closing. Partners with a replicable vertical AI offer in finance, HR, or supply chain are starting to win deals outright.

Three new SAP roles are commanding serious premiums right now and most job boards haven't caught up. AI Solution Architects map how agents, data stores, and governance frameworks fit together in SAP landscapes. Data Engineers for AI build semantic models and vector indexes that feed Joule. Document AI Consultants deploy invoice and document processing systems — high volume, repeatable, and currently the fastest-growing entry point into AI-native SAP work. If you're functional and stagnating, any one of these is a realistic pivot. (IgniteSAP)

Baseline US rates hold firm: 32K average, 1/hr, with S/4HANA architects at 50K+. Contract premiums run 20–40% above permanent. But here's what the salary guides aren't saying: if Klein is right and AI agents automate most routine ERP tasks, the headcount-heavy implementation model — 30 consultants on a 2-year greenfield — is going to compress. The work doesn't disappear, but it concentrates at the top.

The skill that matters most right now: knowing how to govern AI agent outputs in mission-critical ERP processes. A staggering 55% of organizations cite agent hallucination and reliability as their #1 adoption blocker — ahead of cost, privacy, and talent. Consultants who can design validation frameworks for Joule agents in finance or supply chain are walking into a market with almost no competition yet.

KTern.AI on AWS Bedrock — and this one is worth your attention because it's coming directly for SAP consulting hours. KTern.AI has launched agentic AI capabilities on AWS Bedrock that automate three of the most time-consuming phases of any SAP transformation: system discovery, change impact analysis, and test case generation. These aren't generic AI tools layered on top of SAP — they're built specifically for the S/4HANA migration and upgrade lifecycle. (ERP Today)

Real use case: A project team running a brownfield S/4HANA conversion typically spends weeks on business process discovery and impact mapping. KTern.AI's agents compress that into days by automatically scanning the existing landscape, identifying impacted objects, and generating test scripts. That's a direct reduction in the billable hours that fund most transformation project budgets.

Who should care most: Project managers and architects on S/4HANA programs — not because it threatens your role, but because clients are going to start asking why your proposal doesn't include it. Know the tool before your client does.

The question dominating SAP LinkedIn and Reddit right now: "Is SAP still a good career in the AI era?" The optimist case — AI creates new SAP roles faster than it eliminates old ones. The pessimist case — SAP's forward-deployed engineering teams, KTern.AI automating discovery, and 200+ Joule agents are collectively targeting the work that fills most mid-level SAP consultant timesheets. Both camps have evidence. The honest answer: the career is fine if you're moving up the value stack. It's increasingly risky if you're billing hours on work an agent can do.

The ECC migration path debate remains the loudest ongoing argument. Greenfield advocates say you can't build an AI-ready architecture on a legacy process model. Brownfield defenders say greenfield projects blow scope and budget 70% of the time. The emerging third view: use AI-assisted discovery tools (like KTern.AI) to make brownfield conversions faster and cleaner than they used to be — eliminating the greenfield's main argument.

SAP's Developer Challenge "Roll & Move" wrapped July 13–19 on SAP Community. Worth browsing the submissions for practical BTP/CAP implementation patterns. SAP GUI for Windows 8.10 official support also went live July 16 — relevant for teams managing large on-premise estates heading into migrations. (SAP Community)