Cognizant’s Strategic Use of Small Language Models in AI Platforms Q2 2025

Cognizant Technology Solutions Corporation discusses small language models as a key component of its evolving AI strategy, particularly within the broader context of generative and agentic AI platforms. The company positions small language models as part of a differentiated, scalable AI offering that supports enterprise adoption and innovation.

---

Context and Strategic Positioning

1. Integration into AI Platforms and Agentic AI
   Cognizant highlights small language models as foundational elements in building next-generation AI platforms that enable "agentification at scale." These models are embedded within their AI ecosystem, including the Cognizant Agent Foundry, which accelerates enterprise-scale adoption of agentic AI by providing domain-specific small language models, agent templates, and reusable agents.

   "Agent Foundry powered by our Neuro AI suite and a strong AI Partner Ecosystem with hundreds of distinctive agents already available for use. It spans industries and platforms with domain-specific small language models, agent templates and the library of prebuilt agents supporting the full agent life cycle."

2. IP on the Edge and Differentiation
   The company refers to its investment in "IP on the edge," which includes small language models, as a critical differentiator in a fragmented AI market. This "new layer" complements traditional interdisciplinary capabilities by focusing on accuracy, reasoning layers, and responsible AI tooling.

   "To capture the AI opportunity, we recognized and have started building a new layer... This new layer includes translating our award-winning AI labs work to differentiated AI platforms, small language models and tooling to address accuracy of the models, reasoning layers to help make the technology responsible..."

3. Commercial and Pricing Considerations
   Cognizant is actively exploring how to price and package small language models and related AI IP. The company sees this as an opportunity to move beyond traditional productivity gains toward innovation capital that can be monetized in a nonlinear way, creating "sticky" client relationships and premium offerings.

   "We are starting to think about small language models... How do we create the ontology around it and package it into the things we can handover to our project team so that, that becomes real... This is fast evolving. I'm excited about how we could price it, how we could bundle it... Innovation capital, you can monetize in a nonlinear way."

4. Scaling AI Adoption and Talent Development
   The company is scaling innovation by embedding small language models into reusable agents and training its workforce to build AI fluency, supporting rapid deployment and client impact.

   "We believe this is the world's largest initiative of its kind, but what excites me most is the opportunity for every associate to build real AI fluency and deliver continued higher productivity to all our client project work."

---

Business Implications

• Competitive Differentiation: Small language models are part of Cognizant’s strategy to differentiate itself in the AI services market by offering proprietary, scalable AI components that go beyond generic models.
• Client Impact and Adoption: The integration of small language models into agentic AI solutions is driving client adoption and measurable impact, as evidenced by high AI adoption rates and scaling among key accounts.
• Revenue Model Evolution: Cognizant is innovating its pricing and packaging approach around AI IP, including small language models, aiming for premium, outcome-based monetization rather than traditional time-and-materials models.
• Talent and Capability Building: The company is investing heavily in talent development to support the deployment and continuous improvement of AI solutions built on small language models.

---

Summary

Cognizant views small language models as a strategic enabler within its broader AI platform and agentic AI initiatives. These models are integral to creating reusable, domain-specific AI agents that can be rapidly deployed across industries. The company is actively investing in the development, packaging, and commercialization of these models as part of its "IP on the edge" strategy, which it believes will be a key differentiator and growth driver in the evolving AI services landscape.

---

Key quote summarizing the role of small language models:

"This new layer includes... small language models and tooling to address accuracy of the models, reasoning layers to help make the technology responsible... We are starting to think about small language models... how we could price it, how we could bundle it... This is going to be our single biggest differentiator as we take our customers into agentic journeys."