Image generated from Google Gemini

What Is Agentic AI?

My interest for generative Artificial Intelligence (AI) keeps growing. But as I go deeper, I find myself drawn to Agentic AI. This shift raises a few questions, such as what is the difference between generative AI and Agentic AI? And how can process engineering help us understand, structure, and design smart autonomous systems?

A plain LLM solution operates in a reactive mode. It waits for a prompt and then processes it using its language capabilities. Their context is limited to the current interaction. Here, you can think of when you ask a Large Language Model (LLM) chatbot, such as ChatGPT a question.

Moving towards more advanced LLM-based systems, agents leverage the LLM for different sub-tasks. They often operate in response to an initial user prompt and follow a predefined sequence of operations. Here, you can use an LLM-based solution that acts as an intelligent interface, translates a human request and finds relevant information from a knowledge base.

An agentic AI system is designed to understand a high-level goal and independently plan and execute multi-step actions, utilize a diverse set of tools, and adapt its approach based on the outcomes, all aimed at achieving that overarching objective with minimal human intervention.

Some use cases we can find from major vendors are invoice processing, insurance claims processing, compliance workflows, lead follow-up & scheduling in sales, etc.

Once AI agents are implemented, there will be role transformations and the need for new skills. According to a report from McKinsey, workers must learn to collaborate with agents, interpret outputs, and follow ethical standards .

How Agentic AI Impacts Businesses, Employees and Customers

Agentic AI is already transforming business operations and customer experience.

For Businesses: Gartner highlights that agentic AI can transform business decision-making by enabling systems to autonomously initiate tasks based on goals, removing the need for repetitive user interactions. According to Gartner, by 2028, 33% of enterprise software applications will include agentic AI, up from less than 1% in 2024 .

For Employees: Gartner notes that giving AI agency empowers workers to manage tasks via natural language and automation, enabling them to focus on higher-level work .

For Customers: Gartner projects that agentic AI will autonomously handle 80% of customer issues by 2029, enabling proactive resolution without human intervention .

How to Identify Processes Suitable for Agentic AI

Agentic AI is best suited for process steps that are:

  • Goal-driven: Has a clear outcome (e.g., book a meeting, resolve a ticket).

  • Multi-step: Requires sequencing actions (e.g., check eligibility , generate documents, notify user).

  • Tool-interactive: Involves using systems, forms, APIs, or documents.

  • Tolerant to bounded autonomy: Human supervision is minimal or periodic.

  • Routine but dynamic: Involves some judgment or variability, not rigid scripts.

How to Find these Processes

According to McKinsey, “AI initiatives should no longer be scoped around a single use case, but instead around the end-to-end reinvention of a full process or persona journey.” We should adopt a systems view when we rethink workflows, human-computer interactions, and performance metrics.

Business process management and Lean Six Sigma are essential tools when developing agentic AI solutions, especially when aiming to design AI agents that act autonomously, responsibly, and with clear business value. Here’s what we can do:

  • Use Process Mining: Process mining helps identify steps with long wait times, bottlenecks due to manual tasks, repetitive workflows, and other wastes. For example, when we identify process deviations or compliance violations, agents can take corrective action or enforce controls in real time.

  • Identify Automatable Tasks and Decision Points: Process maps reveal where repeatable, structured decisions occur, which is suitable for automation. With Value Stream Mapping (VSM), you can pinpoint repetitive workflows that are ideal for delegation to AI agents trained to act on rules or triggers, and uncover bottlenecks that can be streamlined through agent automation.

  • Understand Dependencies and Guardrails: Mapping provides visibility into what data, systems, approvals, and regulations are involved. You can use process maps to define safe zones where agents can operate autonomously vs. areas requiring human authorization (especially in finance, healthcare, or legal contexts).

  • Prioritize High-Value Use Cases: VSM quantifies time, cost, and impact of each step, so you can focus AI on high-friction or high-cost areas. Using Cost of Poor Quality (COPQ), we calculate the financial impacts of process wastes that can be streamlined with investment in agentic AI, aligned with the Voice of the Customer (VOC) to focus on processes that solve customer irritants.

  • Support Risk Assessment and Control Plans: A mapped process enables Failure Mode and Effects Analysis (FMEA) or other risk tools. With FMEA, you can define failure points if an AI agent acts incorrectly, controls or fallback flows to mitigate those risks and human oversight points to ensure AI agents don’t bypass fraud detection or regulatory checks.

This ensures that agentic AI solutions are not only smart, but aligned with actual operational pain points and outcomes.

Concluding

Agentic AI is being considered the top trend to shape the future with responsible innovation. When deployed successfully, it offers transformative gains in productivity, customer satisfaction, and business agility. But successful implementation requires a blend of process excellence and human-centered design.

This is where a process mindset becomes indispensable. By understanding workflows through techniques like process mapping and value stream analysis, organizations can identify which steps are ready for agentic automation, where human oversight is essential, and how to ensure goal-aligned AI behavior.

The path forward isn’t just about adoption, it’s about transformation. We need to invest in the right foundation to build the knowledge we need to innovate tomorrow.