Palantir in SAP Transformations: The Model, the Momentum, and the Licensing Questions

If Palantir has shown up in an SI proposal on your desk, you are not alone. It is appearing more often in S/4 transformation plans, and the commercial details are not always as clear as the technical ones. In May, we published an UpperEdge post on Palantir’s growing role in SAP transformations and the watch points we advise CIOs to track. In this follow-up, we share what the industry is saying, how Palantir’s model works alongside SAP, what to know about licensing, and our updated view on each watch point. 

What the Industry Is Saying 

SAP made its commitment to Palantir clear at Sapphire in May, presenting it as a strategic partner in its “autonomous enterprise” vision alongside Anthropic, NVIDIA, AWS, and Google Cloud. Palantir AIP is now an SAP Endorsed App, and SAP has announced a dedicated Solution Extension. SAPinsider’s Jehangir Khan describes the relationship as architectural: SAP is treating Palantir as a layer of its own stack. For more on the event, see our Sapphire 2026 Recap: 5 Watchouts Behind SAP’s Announcements. 

The partner ecosystem is growing too. In August 2026, SAP migration specialist SNP announced it is integrating Palantir’s AIP and Ontology into its Kyano platform. The first joint offering automates test data selection, one of the most time-consuming steps in a migration. It is a practical sign that Palantir’s tools are moving into day-to-day delivery work. 

Delivery experience is still building. In June, Khan noted that real-world experience on large SAP-Palantir programs is still “accumulating,” with early adopters learning alongside their providers. For CIOs considering Palantir now, that learning curve belongs in the program timeline and governance from day one. 

There is also a commercial reason for the interest. SAP ties broad access to its Joule AI agents to RISE with SAP. Customers running SAP on-premise, especially on ECC, have more limited access to SAP’s native AI roadmap, and Palantir gives them another route to AI capabilities. For these customers, the choice between Palantir and RISE is a roadmap decision, not just a tooling one. For background, see our Everything We Know about SAP’s AI Strategy. 

How the Model Works 

Palantir’s offering for SAP customers has three layers: 

• Foundry, the data layer. Connects to SAP through an ABAP add-on and combines SAP data with other sources (Palantir cites more than 300) into what it calls a “dynamic digital twin” of the business. It does not change the SAP core. 

• AIP, the AI layer. Adds AI functions and agents that work with that data. 

• Apollo, the deployment layer. Runs Foundry and AIP in cloud, on-premise, or fully disconnected (air-gapped) environments, which matters for regulated industries. 

SAP remains the system of record. Palantir describes its role alongside SAP as a “system of action,” and on S/4 programs it is most often positioned for data migration, testing support, and operations after go-live. Central to that positioning is what Palantir calls the ontology: a model of how your business runs, built and maintained inside the platform. Its value on S/4 programs has yet to be demonstrated at scale, but its design has a clear implication for buyers. The more business logic that lives in the ontology, the harder it becomes to leave. That is why we advise addressing data ownership and exit terms early: moving away later means rebuilding that business logic, not just migrating data. 

Licensing: What CIOs Should Know 

There is no public price list for Foundry or AIP. Fees are negotiated deal by deal and grow with the scope of the deployment. That means pricing can vary between similar programs, and the Palantir negotiation deserves the same attention as the SI negotiation. 

Clarity starts in the SI proposal. In the SI proposals we evaluate for clients, Palantir often appears in the solution architecture without clear treatment of its licensing in the commercial section. That makes it hard to tell whether the SI’s price includes Palantir or whether licensing will be contracted separately, and it makes bids harder to compare like-for-like. The simplest fix is to have each bidder show Palantir licensing and FDE costs as separate lines. 

The subscription is only part of the cost. Public estimates vary widely with scope, so treat any figure you hear as a starting point, not a benchmark. Implementation and the people you will need to run the platform day to day add to the total. Clients should also confirm how many forward-deployed engineers (FDEs), Palantir’s embedded engineers, the deal assumes and whether they are included in the subscription. 

Agree on terms before a pilot starts. We are seeing SI proposals that place a Palantir proof of concept at the front of the plan, sometimes funded by the SI as an investment. A pilot is a sensible way to test fit. It is also the point where you have the most negotiating leverage, so we recommend agreeing on the key commercial terms for the full program before it begins.  

Where May’s Watch Points Stand Now 

Our prior post flagged four watch points and one upside case. Most are still open, and two of them, proposal transparency and commercial accountability, need the most attention from buyers. Here is our updated view on each. 

• Palantir works through partners: unchanged. Palantir continues to be delivered with SIs and specialist partners, as SNP’s August announcement shows. The lead SI remains the main point of accountability, and the contract should reflect that. 

• Reference programs: still building. Delivery experience on large SAP-Palantir programs is still accumulating. Palantir and its SI partners have yet to share a completed, referenceable SAP transformation that shows Palantir’s measured impact, even when clients ask for one. We expect that to change as early programs go live. 

• Proposal transparency: still a gap. SI proposals still do not give a complete commercial picture of Palantir’s role. Licensing, FDE costs, the contracting model, and the efficiency claims behind the price are often unclear or missing. Clients should be able to see all four in the proposal, and they are among the first things we look for when we review one. 

• Commercial accountability: still unresolved, and the biggest issue for buyers. This is where clients are struggling most. SIs and Palantir present significant efficiency gains in their proposals, but when clients ask them to stand behind those gains commercially, through fees at risk or contractual timeline and cost commitments, the commitments rarely follow, in part because there are not yet enough completed programs to back them. That leaves the client carrying the risk for benefits it is being asked to pay for. If faster timelines or lower costs are part of the case for Palantir, SIs and Palantir should be prepared to commit to them in the contract. If they will not, CIOs should treat the claimed savings as unproven and weigh the proposal accordingly. 

• The upside case: promising, not yet proven. SNP, Accenture, and Palantir continue to point to significantly shorter timelines, in some cases by as much as half. Until completed programs confirm those results, the prudent approach is to treat them as a target to validate, not a baseline for the business case. 

Our Advice for CIOs 

The direction is clear: SAP is building Palantir into its AI strategy, the partner ecosystem is growing, and the technology is maturing. Pricing and commercial structures are still taking shape, so the clarity CIOs need will come from pressing SIs and Palantir for commercial transparency and commitments during evaluation and negotiation, not from waiting for the market to settle. 

If you are running a competitive RFP, expect SIs to bring different levels of Palantir experience. Accenture and Deloitte have announced Palantir alliances. Other SIs may propose Palantir when asked or offer their own AI and data platforms. Either can be the right fit. The useful question for each bidder is how they would use Palantir, what experience they bring, and what they would propose instead. 

Questions to Ask Before You Sign 

• Is Palantir licensing included in the SI’s price, or will it be contracted separately? 

• Which commercial terms for the full program will we agree on before a pilot starts, such as pricing beyond the pilot, limits on increases at renewal, and our rights to export our data? 

• If we stop using Palantir later, how would we export our data and ontology, and what would that involve? 

• Does our budget cover implementation and internal staffing, not just the subscription? 

• If expected timeline savings or platform performance fall short, how does the contract share accountability between the SI and Palantir? 

• How does the SI’s delivery approach, staffing, and plan change when Palantir is in scope, and where is that reflected in the proposal? 

• Have we compared Palantir with SAP’s own AI options, including what a RISE commitment would provide? 

• What Palantir experience does the SI bring, and how does Palantir compare with the SI’s alternative on cost and expected value? 

These questions are a starting point. Asking them is the easy part. Knowing what a good answer looks like takes current market insight and experience across many SI and software negotiations. As an independent advisor, that is what we bring, in the context of your own proposal, program, and negotiating position.  

If you are planning an S/4 transformation and want a second set of eyes on your proposal, we are glad to help. Request a call to discuss what we can do for you.