Rama Akkiraju de Nvidia on how architects of the AI platform help reject the business vision and technical execution
Companies in all industries explore AI to rethink problem solving and redefine business processes. But the success of these companies requires the right infrastructure, such as AI factories, which allow companies to convert data into chips and results.
Rama Akkiraju, vice-president of it for AI and automatic learning at Nvidia, joined the AI podcast to discuss how companies can build the right foundations for the success of the AI.
Based on more than two decades of experience in the field, Akkiraju has provided its point of view on the evolution of AI, from the perception of AI to AI Generative to an agentic AI, which allows systems to reason, plan and act independently, as well as HAVEwhich allows autonomous machines to act in the real world.
What is striking, underlined Akkiraju is the acceleration of the evolution of technology: the transition from the perception to the generating AI took around 30 years, but the jump to agency AI occurred in just two. She also pointed out that the AI transforms the development of software by becoming an integral layer in the architecture of applications – not just a tool.
“Treat AI as a new layer in the development stack, which fundamentally reshapes the way we write software,” she said.
Akkiraju also talked about the essential role of architects of the AI platform in the design and construction of AI infrastructure based on specific commercial needs. Company implementations require complex batteries, including data ingestion pipelines, vector databases, security controls and evaluation frames – and platform architects are used as a bridge between strategic commercial vision and technical execution.
In the future, Akkiraju has identified three trends shaping the future of AI infrastructure: the integration of AI architecture specializing in native business systems, the emergence of models specific to the field and optimized equipment for specific use cases, and increasingly autonomous agent systems requiring sophisticated memory and context management.
Hourly
1:27 – How the Akkiraju team builds platforms, chatbots and corporate co -pilotes.
4:49 – The accelerated evolution of the perception of AI at generative AI to agency AI.
11:22 – The complete battery required to implement AI in corporate settings.
29:53 – Three major trends sharing the future of IA infrastructure.
You might also like …
Jacob Liberman of Nvidia on the creation of agentics to companies
Jacob Liberman, director of product management at NVIDIA, explains how agental AIs fill the gap between powerful AI models and practical business applications, allowing intelligent multi-agent systems that reason, act and perform complex tasks with autonomy.
Isomorphic laboratories rethink the discovery of drugs with AI
The leadership team of isomorphic laboratories discusses their AI-First approach for the discovery of drugs, considering biology as an information processing system and creating generalizable AI models capable of learning from the whole universe protein and chemical interactions.
AI agents pass digital experiences at the higher level of the game and beyond
AI agents with advanced perception and cognition capacities make digital experiences more dynamic and personalized in all industries. Chris Covert of Inworld AI explains how intelligent digital humans reshape interactive experiences, from health care games.
