IFS launched IFS Intelligence at its Unleashed 2026 conference in Orlando, Florida, this week, offering industrial customers an AI platform that connects conversational tools and agent workflows to company operational data. CEO Mark Moffat said the platform is aimed at industries including aerospace, energy and manufacturing, where a wrong answer can ground an aircraft, stop a flight or threaten human life. IFS says its customers’ combined asset intelligence covers more than 2.4 trillion items.

Moffat contrasted enterprise data with AI model training data, saying the latter amounts to a petabyte, equivalent to 50 hard drives and small enough to fit in an aircraft carry-on. The article notes that clean training data may be compressed from hundreds of petabytes of raw data. Moffat said models lack company-specific knowledge, such as gearbox warnings recorded a month before a breakdown and 20 years of work orders, fault codes and repairs for individual assets. Some employee knowledge has never been written down.

IFS Intelligence is available now under a universal enterprise-wide license, with AI execution subject to a fair-use cap. It works across relevant IFS software and can connect to other systems. It draws on live and historical data about assets, work orders, schedules and service histories to report current conditions, propose next steps and carry out plans after approval. Its Cowork interface accepts chat or voice requests, can be reached through MCP from connected systems such as Microsoft 365, and lets users save live data views as “Lenses” and successful actions as repeatable “Skills.”

The platform’s Loops component coordinates agent actions for tasks such as replenishing materials, preparing customer orders and drafting service plans. Based on technology from IFS’s acquisition of agentic AI startup TheLoops last year, its digital workers refer decisions requiring judgment or approval to people. Customers can configure ready-made workflows or create their own in a no-code Agent Studio. Runway turns plain-language requests to configure, extend or migrate IFS software into specifications and working prototypes.

The Industrial AI Harness supplies operational context, permissions, governance and controls, inheriting user identities and access rules from IFS Cloud and connected systems. Its functions include planning and coordinating agents, managing skills and system connectors, carrying task context forward, enforcing approval limits, checking work against policies, escalating when needed, and logging decisions and actions for review and audit.

IFS also launched Fulfillment Flow for its IFS Softeon warehouse management platform, acquired through the purchase of 7bridges last year. It uses real-time warehouse context to coordinate six AI agents focused on different parts of operations; people set priorities and make final decisions on proposed actions. Kodiak Gas Services CIO Pedro Buhigas said the company achieved an 85% efficiency gain after deploying a Loops workflow for materials replenishment, then introduced other workflows. He said he hoped to return the following year to discuss hundreds of agents deployed across the organization.

IFS launched IFS Intelligence at its Unleashed 2026 conference in Orlando, Florida, this week, offering industrial customers an AI platform that connects conversational tools and agent workflows to company operational data. CEO Mark Moffat said the platform is aimed at industries including aerospace, energy and manufacturing, where a wrong answer can ground an aircraft, stop a flight or threaten human life. IFS says its customers’ combined asset intelligence covers more than 2.4 trillion items.

Moffat contrasted enterprise data with AI model training data, saying the latter amounts to a petabyte, equivalent to 50 hard drives and small enough to fit in an aircraft carry-on. The article notes that clean training data may be compressed from hundreds of petabytes of raw data. Moffat said models lack company-specific knowledge, such as gearbox warnings recorded a month before a breakdown and 20 years of work orders, fault codes and repairs for individual assets. Some employee knowledge has never been written down.

IFS Intelligence is available now under a universal enterprise-wide license, with AI execution subject to a fair-use cap. It works across relevant IFS software and can connect to other systems. It draws on live and historical data about assets, work orders, schedules and service histories to report current conditions, propose next steps and carry out plans after approval. Its Cowork interface accepts chat or voice requests, can be reached through MCP from connected systems such as Microsoft 365, and lets users save live data views as “Lenses” and successful actions as repeatable “Skills.”

The platform’s Loops component coordinates agent actions for tasks such as replenishing materials, preparing customer orders and drafting service plans. Based on technology from IFS’s acquisition of agentic AI startup TheLoops last year, its digital workers refer decisions requiring judgment or approval to people. Customers can configure ready-made workflows or create their own in a no-code Agent Studio. Runway turns plain-language requests to configure, extend or migrate IFS software into specifications and working prototypes.

The Industrial AI Harness supplies operational context, permissions, governance and controls, inheriting user identities and access rules from IFS Cloud and connected systems. Its functions include planning and coordinating agents, managing skills and system connectors, carrying task context forward, enforcing approval limits, checking work against policies, escalating when needed, and logging decisions and actions for review and audit.

IFS also launched Fulfillment Flow for its IFS Softeon warehouse management platform, acquired through the purchase of 7bridges last year. It uses real-time warehouse context to coordinate six AI agents focused on different parts of operations; people set priorities and make final decisions on proposed actions. Kodiak Gas Services CIO Pedro Buhigas said the company achieved an 85% efficiency gain after deploying a Loops workflow for materials replenishment, then introduced other workflows. He said he hoped to return the following year to discuss hundreds of agents deployed across the organization.