Our CTO of AI Babak Hodjat kicked things off at the Newsweek AI Impact Summit today and shared his thinking on agentic AI:
🚀 The year of agents: a new era in AI
At Cognizant’s R&D Lab and AI Studio, we’re not just advancing core AI research—we’re shaping the future of intelligent systems. From building being the foremost lab for AI systems and decision making to spearheading “AI for Good” our mission is clear: innovation with impact.
🔍 Why agents are different
Unlike generative AI models that simply transform input into output, **agents are empowered to act**. They’re knowledge workers in a box—engineered by humans to perform tasks, retrieve data, write code, and collaborate with other agents.
🧠 The power of multi-agent systems
We’re building networks of agents that coordinate in natural language, share context, and dynamically adapt to new business needs. Each agent contributes to a shared canvas, combining model intelligence with executable code.
💡 Design principles
An agent is both a model (a black box) and code (fully controllable). This duality lets us engineer autonomy, define job descriptions, and ensure ethical alignment. Success isn’t just about performance—it’s about responsibility.
🌐 Interoperability & evolution
These systems aren’t static. They evolve, spawn, and merge on the fly. Whether it’s travel decisions or cost estimators, agentic systems will soon be part of everyday consumer experiences.
👥 Collaboration & oversight
We work with partners to discover processes, build agent nuclei, and continuously refine them. Review mechanisms ensure agents meet standards for success rate, cost, response time, and compliance.
This is more than a trend—it’s a transformation.
Welcome to the age of agents.