Female doctor looking at patient undergoing CT scan. Doctor in uniform using tomographic machine with … [+]
HE is nervous. We know that a certain amount of Artificial Intelligence is somewhat nervous (or perhaps nervous) in its implementation due to concerns over decision-making biases, questions over the origin of the data sets and the underlying models from which it is composed… and because of the environments it’s allowed to operate in and whether or not it has guardrails in the back to make sure it does what we humans intended.
But AI is also “neural” in the sense that it can now be classified as neural-powered, with neural networks being those with enough logical computational power to mimic the human brain, particularly in areas such as image recognition, image processing, natural language and decisions. -making – the latter factor is as important in both industrial enterprise applications as it is in games, as we seek to interact with console characters that appear as human as possible.
Cognizant Cognoscenti
The work to provide neuro-acceleration in the AI space is cognitive. Essentially a global professional services company (you could say a business process outsourcing company if you were less polite), Cognizant has around 350,000 technology professionals working on digital transformation projects across industries, including life sciences, communications, financial services and the automotive sector.
The company’s latest development of its Neuro AI Multi-Agent Accelerator and multi-agent service suite (not known as NAIMAMASS, but we wish it was) is designed to help organizations accelerate the development of AI agents. As we know so far, AI agents are dedicated iterations of AI functionality that operate increasingly autonomously without human intervention towards defined goals as they learn as they go, so that they become more adept at tasks theirs over time.
When are AI agents useful?
To further clarify, AI agents can be particularly useful when applied to workplace tasks that are characterized by the need for adaptive operations, real-time decision making, and user personalization.
“AI agents are transforming enterprise operations by automating tasks and reducing manual effort, enabling employees to focus on strategic activities,” said Babak Hodjat, CTO of AI at Cognizant. “However, without collaboration between specialized agents, software systems will remain disconnected from larger business goals. The Neuro AI Multi-Agent Accelerator and Multi-Agent Service Pack allow customers to build and deploy agents across their organization’s fabric.” them, so they can work together across businesses to help people in many roles, from finance and IT to marketing and sales.”
The company says that traditional workflows and “fixed automation” (ie, software intelligence functions from robotic process automation to any other form of pre-architected shortcuts including the most basic forms of AI) no longer meet customer expectations for due to rising costs and demand. for real-time adaptability.
As we know from the current hype cycle, AI agent systems are helping to bridge the gap that simple old-fashioned fixed-wheel automation offered us.
Prebuilds without code
Designed to accelerate the adoption of agent AI, Cognizant’s neuro AI tool can be described as a code-free development framework that includes a collection of pre-built reference agent networks to enable businesses to rapidly prototype, customize and scale multi-agent systems. Pre-built multi-agent network models apply to industry-specific processes, including supply chain management, customer service and insurance underwriting. Additional agent networks can be rapidly created using natural language descriptions to suit different scenarios and customer use cases. They can also be extended to include third-party agents.
According to Hoxhat and team, “Multi-agent systems go beyond single agents by enabling decentralized decision-making, where agents act independently but cooperatively to solve complex, interdependent problems. They are designed to provide scalability across functions and geographies, allow expansion without overhauling systems, and provide resilience through redundancy, ensuring continuity even if individual agents fail.
Agents on a mission
As we now work to spread the use of agent AI across the enterprise, Cognizant will hope that its technology will be leveraged to refine the use of autonomous agent networks in enterprise workflows across industries. It seeks to provide a structured framework for agents while enabling interaction and coordination between multiple agents using a standardized and tested approach.
“We are entering a pivotal era for human-plus-machine collaboration. Enterprises that continue with independent agents to duplicate human labor will struggle to achieve value,” said Phil Fersht, CEO and principal analyst, HFS Research. “Rather, they must deploy true organizational intelligence where agents and their human counterparts bring contextual intelligence to improve workflows, data sets and processes to deliver results.”
Cognizant provides the ability to deploy new multi-agent networks that can be rapidly created using natural language descriptions to suit different scenarios and customer use cases. It can integrate newly developed multi-agent networks with pre-existing and third-party agent systems using application programming interfaces and enable agent responsibilities summarization that allows tasks to be scoped and automatically routed to the appropriate AI agents . We also find solutions to ambiguity where a network of agents can exist, which allows new agents to be added more easily minimizing errors and improving response time… and (fortunately) this technology can handle large workloads distributing tasks across multiple servers.
As agentic artificial intelligence comes into play now, accelerated actions of agency (this is not a word accepted in all dictionaries, but it is a term we may need to get used to) will become part of the fabric of how this AI race has been brought to life. IT shelves of all faiths and beliefs.
