First it was generative AI, then artificial general intelligence. Now, the spotlight is on agentic AI – autonomous software agents capable of completing tasks like scheduling meetings, handling customer service, and even writing code.
Unlike traditional AI assistants that require prompts to operate within narrow guardrails, agentic AI agents can pursue goals, make decisions, and take action in real-time business environments.
For executives looking to automate complex workflows and reduce manual labor, the appeal is obvious.
The funding surge
Venture capital firms are actively investing in agentic AI startups. For example, Cognition, the team behind the AI coding assistant Devin, raised $175 million at a $2 billion valuation in one of the space’s largest funding rounds.
TechCrunch reported that Adept AI, which is developing general-purpose action agents for desktops, pulled in over $350 million in Series B funding.
These large early investments signal that VCs believe agentic AI isn’t just the next hype cycle, but a foundational shift in enterprise technology.
The enterprise adoption problem
Despite the buzz, most companies remain stuck in pilot mode. According to Accenture, just 36% of executives say their organizations have scaled generative AI solutions, and only 13% report achieving significant enterprise-level impact. Meanwhile, Gartner projects agentic AI will power one-third of enterprise software by 2028, up from just 1% in 2024.
So what’s holding it back?
“Many leaders underestimate the time, effort, and resources required for successful integration,” said Aishwarya Singh, SVP of Digital Collaboration Services at NTT Data.
Her team launched a full-stack agentic AI offering earlier this year to address exactly that challenge. Built with Microsoft Copilot Studio and Azure AI Foundry, NTT’s system supports the full lifecycle: advisory, build, implementation, monitoring, retraining and optimization.
Infrastructure and talent: the missing links
In an interview with Forbes, Singh noted “The biggest economic bottlenecks include the high initial investment in infrastructure and technology, the cost of integrating AI with existing systems and the need for specialized talent to manage and maintain AI systems.”
For those with systems in place, agentic AI can deliver. NTT Data saw productivity gains of 50–65% in internal ticketing systems during early trials. By designing multi-agent workflows connected through omnichannel large language models like voice, email, and chat, the company was able to automate complex service processes.
But success requires more than deployment. It likely requires performance tracking, security audits, alignment with business goals, and continuous retraining.
The bottom line
The agentic AI wave is real, and the funding surge shows investors are betting on a fundamental shift. But for it to be a success, leaders will likely be those who don’t just build smarter agents, but also engineer the full ecosystem needed to support and scale them.
German 
English