Agentic AI could create nearly $450 billion in value by 2028, helping organizations automate and orchestrate workflows if they reimagine their processes.
Right now, only 2% of organizations have fully scaled agentic AI deployments, according to Capgemini Research Institute data. Part of the challenge many organizations face is the approach they take: They add AI agents to existing processes, rather than reimagine those processes entirely. By using a zero-based process redesign (ZBPR) strategy instead, organizations can leverage the full power of agentic AI to achieve operational efficiency at scale.
What is ZBPR, and why is it so important for agentic AI scalability? ZBPR requires remodeling processes from the ground up based on what agents can do, not the way things were done before. ZBPR avoids automating suboptimal workflows, and it makes the most of AI's agentic and orchestration capabilities. Designing an agent-native process gives organizations the opportunity to eliminate steps that don't create value or support compliance, and a chance to repurpose manual labor to better align the process align with strategic business goals.
Related:Okta's Harish Peri on what it takes for CIOs to secure AI agents
Rather than deploying agentic AI incrementally within existing processes, ZBPR harnesses the contextual awareness, reasoning, planning and acting capabilities of agentic AI to radically transform processes. The result is workflows that optimize costs and efficiency, support risk management and compliance, and are more scalable and flexible than legacy processes without linear cost increases.
For example, an employee onboarding process that used to require signing in to multiple platforms to handle account creation, payroll and equipment could be redesigned using a team of orchestrated agents. Those agents could handle all onboarding tasks for each new hire through a single point of contact. The agents and process could be adapted across different geographies and acquisitions for flexible scalability.
Other benefits of agent-native processes include the ability to run around the clock with minimal human oversight, shorter cycle times, more accurate data entry and real-time data visibility for compliance and insights. By eliminating rote, repetitive tasks, ZBPR can also increase employee availability for higher-value, more engaging work, such as dealing with edge cases and developing strategies based on agentic process insights. For example, with an expense-processing agent handling all employee expense reports below a certain threshold of value, the human manager can focus on higher-value reports and those with flagged anomalies.
Related:Intuit's chief AI officer on the SaaSpocalypse and disciplined AI
Organizations that have already implemented ZBPR for their agentic process transformation report higher AI ROI than organizations using less holistic automation strategies. The ROI improvement is due to not just greater efficiency. In some cases, the ZBPR plus AI approach allows organizations to automate workflows that previously couldn't be automated.
Consider an insurance contact center, where policyholders call or message with many different types of claims, varying levels of coverage and a patchwork of state laws governing their policies. An incremental agentic strategy would use a chatbot to handle the most basic inquiries, saving some time but not fundamentally transforming the experience for policyholders or service agents. A ZBPR redesign of the contact center workflow can fully leverage agentic capabilities.
For example, one general AI agent can triage and process most basic contacts from end to end. A team of more specialized agents can handle a large portion of the contacts that the main agent isn't trained to deal with. That leaves a smaller set of more complex or high-value issues for human agents to handle. By using ZBPR, the contact center can reduce costs, resolve issues faster for better customer experience and allow human agents to focus on the areas where their judgement and empathy matter the most.
Related:Time for an AI exit strategy: How CIOs are cutting AI waste
This example highlights a key trend. The most innovative adopters of agentic AI are shifting away from task-level automation to build multi-agent, end-to-end workflows that deliver more value than automating individual tasks. This shift is critical for organizations that want to future-proof for efficiency, agility and resilience.
Adopting a zero-based process redesign mindset requires a shift that starts at the top. Executives need to develop and share a clear vision of what's possible with agents and identify high-impact processes to pilot this approach. Next, zero-based process design workshops create agent-native workflows that achieve process outcomes more efficiently and support business goals for value creation.
For each redesigned workflow, the organization must orchestrate multi-agent teams to handle all relevant processes. Humans must be in the loop as safeguards for edge cases and for compliance monitoring using real-time process data. As agentic pilots scale, organizations will need to reskill growing numbers of employees to manage AI agents or end-to-end agentic workflows. Reskilling should be part of a larger, ongoing cultural shift that positions agentic automation as a way to elevate employees' capabilities rather than replace them. Successful ZBPR transformations will depend heavily on compliance, governance and change management to ensure that employees and agents work together.
As organizations build out agentic workflows, change their culture and reskill their employees, they may benefit from creating a center of excellence for automation that tracks value at each step of the agentic transformation. A center of excellence can also help develop the next iteration of agentic AI value creation, whatever form that may take. For now, however, the key fact is that the future of processes and workflows isn't simply AI-assisted: It's AI-orchestrated and largely self-managing if organizations are bold enough to reimagine the way they work.
Capgemini Invent
Jan Malte Prädel is a senior director at Capgemini Invent in New York. Connecting insights from business process analysis and process data mining, he helps organizations uncover, analyze and solve business execution gaps.
Jan Malte has more than 20 years of experience in business process and IT consulting across North America and Europe, spanning diverse sectors and process domains. He specializes in helping organizations build programs, teams and momentum to tackle business challenges in a data-driven way that overcomes resistance toward successful transformation.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。