Palantir (PLTR 0.30%) has already shown what it takes to be a successful enterprise artificial intelligence (AI) player: Become the core platform for customers to build their AI applications on, rapidly turn pilot projects into production-level deployments, cross-sell and upsell to existing clients, and focus on new client acquisition across industries and new verticals.
Image source: Getty Images
Innodata (INOD 0.57%) is much smaller, but it seems to be on a similar growth trajectory. The company is moving beyond traditional data services and is now becoming an AI partner focused on the data and evaluation layer in the enterprise AI stack -- something that Palantir is not focusing on.
Financial performance
Palantir's second-quarter fiscal 2025 (ending June 30) earnings performance underscores the success of this business model. Revenues grew 48% year over year to over $1 billion, with U.S. commercial and U.S. government revenues soaring year over year by 93% and 53%, respectively. The company's Rule of 40 score increased 11 percentage points sequentially to 94. Management raised its fiscal 2025 revenue guidance and ended Q2 with total contract value (TCV) of $2.3 billion.
NASDAQ: INOD
Key Data Points
Innodata's Q2 of fiscal 2025 (ending June 30) performance was also stellar. Revenues grew 79% year over year to $58.4 million, while adjusted EBITDA increased 375% to $13.2 million. Management raised full-year organic growth guidance to 45% or more, driven by a robust project pipeline, with several projects from large customers.
Data vendor to AI partner
Palantir differs from other AI giants by focusing not on large language models, but on its ability to leverage AI capabilities to resolve real-world problems. The company's focus on ontology -- a framework relating the company's real assets to digital assets -- helps its software properly understand context to deliver effective results.
Innodata also seems to be implementing a similar strategy. Instead of focusing on traditional data and workflows, it is providing "smart data," or high-quality complex training data, to improve accuracy, safety, coherence, and reasoning in AI models of enterprise clients. It is also working closely with big technology customers to test models, find performance gaps, and deliver the data and evaluation needed to raise model performance. That shift will help Innodata's offerings become entrenched in their clients' ecosystems, thereby strengthening pricing power and creating a sticky customer base.
Vendor neutrality
Palantir has not built any proprietary foundational model. Plus, its Foundry and artificial intelligence platform (AIP) can run on any cloud and can be integrated with multiple large language models. By giving its clients the flexibility to choose their preferred cloud infrastructure and AI models, the company prevents vendor lock-in. This vendor neutrality has helped build trust among both government and commercial clients.
Innodata's vendor-neutral stance is also becoming a competitive advantage. In its Q2 earnings call, an analyst noted that several big technology companies have said they would no longer work with Innodata's largest competitor, Scale AI, after Meta Platforms' large investment in the company. This is creating new opportunities for Innodata. Because it isn't tied to any single platform, there is no conflict of interest involved in working with Innodata. This gives enterprises and hyperscalers confidence that their proprietary data and model development efforts will not be compromised.
Scaling efforts
Palantir's business is seeing rapid traction, driven primarily by high-value clients. The company closed 157 deals worth $1 million or more, of which 42 deals were worth $10 million or more.
Innodata is scaling up revenues while also focusing on profitability. Management highlighted that it has won several new projects from its largest customer. The company has also expanded revenues from another big technology client, from $200,000 over the past year to an expected $10 million in the second half of 2025. Innodata's adjusted EBITDA margins were 23% in the second quarter, up from 9% the same quarter of the prior year.
Agentic AI
Palantir has been focusing on the agentic AI opportunity by investing in AI Function-Driven Engineering (FDE) capability within its AIP platform. AI FDE is expected to solve bigger and more complex problems for clients by autonomously executing a wide array of tasks, including building and changing ontology, building data flows, writing functions, fixing errors, and building applications. It also works in collaboration with humans and can help clients get results faster. Palantir is thus progressing toward developing AI systems that can plan, act, and improve inside enterprise setups.
Innodata is also advancing its agentic AI capabilities by helping enterprises build and manage AI that can act autonomously. The company aims to provide simulation training data to show how humans solve complex problems, and advanced trust and safety monitoring to guide these systems. Agentic AI is also expected to help the robotics field progress rapidly, and AI systems will run on edge devices used in daily life. Hence, Innodata plans to invest more in building data and evaluation services for these agentic AI and robotics projects, which it expects could become a market even larger than today's post-training data work.
Valuation
Despite its many strengths, Innodata is still very much in the early stages of its AI journey. Shares have gained by over 315% in the last year. Yet, with a market cap of about $1.9 billion and trading at nearly 8.2 times sales, Innodata is priced like a data services company making inroads in the AI market, and not like an AI platform company with a significant competitive moat. On the other hand, Palantir stock is expensive and trades closer to 114 times sales. This shows how Wall Street rewards a category leader like Palantir, whose offerings act as an operating layer for enterprise AI companies.
Innodata also needs to dominate the AI performance market to reach such sky-high valuations. The company will need to expand its customer base, cross-sell and upsell to existing clients, and make it difficult to switch to the competition.
While this involves significant execution risk, there is definitely a chance -- albeit a small one -- that Innodata can become the next Palantir in the 2030s.
