Every AI company in the world is trying to remove humans from the equation. Palantir is putting them back in and charging a premium for it.
Wedbush attended Palantir's tenth AIPCon this week, and the takeaway from analyst Dan Ives was not about the platform's technical capabilities. It was about a job title: the Forward Deployed Engineer.
Palantir's FDEs sit alongside customers in the field. They listen to the problem. They build the solution in real time while the customer watches. The customer sees changes happen immediately, understands what the software is doing, and trusts it because a human built it in front of them.
In an industry obsessed with autonomous agents and zero-touch automation, Palantir is winning deals because it sends a person to your office.
What sits underneath
The human layer sits on top of something genuinely powerful. Palantir's ontology, the semantic model that represents the nouns and verbs of a business, allows AIP to contextualise data at machine speed and grow with the organisation. Every decision builds on the last. The compounding value deepens the integration. The switching costs rise with every deployment.
Wedbush described it as synthesising vast amounts of data across entire organisations in minutes, producing conclusions that would otherwise take professionals days or weeks. The platform does not replace humans. It puts all the information in front of them and lets them decide.
That distinction is commercially important. Enterprises in regulated industries, defence, insurance, construction, energy, are not ready to hand decisions to an AI. They want AI that makes their people faster and better informed. Palantir sells exactly that.
Partnership expansion
AIPCon 10 showcased a broadening partner ecosystem. Palantir announced a multi-tiered partnership with Google Cloud, making the platform available on Google Cloud Marketplace and enabling two-way data federation between BigQuery and Foundry, semantic exchange between Google's Knowledge Catalog and the Ontology, and connectivity between Gemini and AIP.
The Google deal is significant because it lets Palantir customers plug the best available AI models into their workflows without being locked into a single provider. The ontology becomes the stable layer. The models are interchangeable.
Other announcements included an enterprise expansion with Mexican insurer GNP Seguros for underwriting, claims processing and fraud prevention, and a multi-year strategic partnership with construction firm McCarthy to build an AI operating system spanning 160 years of construction intelligence.
Wedbush also noted conversations with a copper importer using Palantir to optimise refinement throughput, a private jet management business generating more precise quotes by combining public and proprietary data, and a data centre operator using AIP to monitor grid performance.
The valuation problem that will not go away
Palantir trades at $141.70, giving it a market capitalisation of $341bn. Wedbush maintains an outperform rating with a $230 price target, implying 62% upside. The stock trades at roughly 84 times forward earnings.
That multiple prices in sustained hyper-growth, expanding margins and continued enterprise adoption at a pace that justifies one of the highest valuations in the software sector. CrowdStrike and Datadog, both growing rapidly, trade at lower multiples.
The bull case is that Palantir's ontology creates a compounding moat that no competitor can replicate, and that the FDE model drives land-and-expand adoption that accelerates over time. The bear case is that an 84x earnings multiple for a company growing revenue at 45% requires perfection, and perfection in enterprise software is rare.
Contrarian insight
The most interesting thing about Palantir is not the AI. It is the business model that wraps around it.
Every other AI company is trying to scale by removing humans from the process. Palantir scales by deploying humans who make the AI useful. The FDE model is expensive, labour-intensive and impossible to automate, which is precisely why it works and why competitors cannot copy it.
In a market where every enterprise software company claims to be AI-native, Palantir's differentiator is that it sends a human being to sit next to you and solve your problem while you watch. The AI does the heavy lifting. The human does the selling.
It is the least scalable approach in the industry. It is also the stickiest. And at $341bn, the market is betting that sticky wins.
