Driving better business outcomes through multi-agent systems

2025-09-04 · Source: AIOps Blog – BMC Software | Blogs · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Operations & Process Management · Depth: Intermediate, short

Summary

Agentic AI is increasingly mandated in enterprises to convert data into actionable knowledge and boost productivity, despite a Gartner prediction that over 40% of projects will be canceled by late 2027 due to cost, unclear value, or inadequate risk controls. This technology involves designing systems for business value, not just plug-and-play solutions. Agentic AI differs from traditional automation by using iterative planning to autonomously solve complex, multi-step problems, exemplified by non-deterministic agents like BMC HelixGPT Situation Observer for dynamic analysis and deterministic agents such as BMC HelixGPT Employee Navigator for pre-determined workflows. A multi-agent system approach, like BMC HelixGPT Best Action Planner, orchestrates various models, tools, and data to diagnose and resolve issues by processing logs, metrics, traces, and change data, mirroring human SRE/DevOps processes.

Key takeaway

For CIOs and IT leaders evaluating agentic AI, recognize that successful adoption hinges on a strategic multi-agent system approach rather than isolated agent deployments. Focus on rethinking entire workflows and aligning agentic AI with strategic goals like revenue growth, cost optimization, and risk management to avoid project cancellations and realize substantial enterprise productivity gains.

Key insights

Successful agentic AI implementation requires strategic multi-agent systems, not isolated agents, to drive enterprise productivity.

Principles

• Prioritize business value over technology for agentic AI.
• Orchestrate multi-agent systems for complex problem-solving.
• Rethink workflows entirely for AI transformation.

Method

Implement a multi-agent framework from the outset, breaking down AI agents into sub-agents dictated by new workflows, and asking critical questions about safe decision offloading and action standardization.

In practice

• Use non-deterministic agents for dynamic context analysis.
• Employ deterministic agents for structured self-service support.
• Integrate multi-agent systems for incident resolution.

Topics

• Agentic AI
• Multi-Agent Systems
• IT Operations
• Workflow Automation
• Enterprise Productivity

Best for: CTO, VP of Engineering/Data, Executive, Director of AI/ML, MLOps Engineer, AI Architect

Related on AIssential

• From Hype To Outcomes: How VCs Recalibrate Around Agentic AI — Agentic AI is shifting from ambitious autonomy to practical, outcome-driven applications with human oversight.
• The Internet of Agents and What It Means for Enterprise Leaders - with Vijoy Pandey of Outshift by Cisco — Enterprises are transitioning to probabilistic, agent-driven AI, necessitating new approaches to interoperability, governance, and access control.
• How Autonomous AI Agents Are Reshaping the AI Game of Productivity — Autonomous AI agents enhance productivity by independently managing tasks, freeing human focus for creative and strategic work.
• How AI Agents Can Automate Complex Tasks — AI agents automate complex, multi-step tasks by dynamically planning, reasoning, and adapting with minimal human oversight.
• 10 leading enterprises show why agents mean business — Enterprises are deploying autonomous AI agents to re-engineer operations, automating complex tasks and scaling business functions.
• Scaling Enterprise Productivity Through Agentic AI and Workflow Automation - with Kuo Zhang of Alibaba.com — Agentic AI is democratizing global sourcing and enhancing enterprise productivity by automating complex workflows.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AIOps Blog – BMC Software | Blogs.