+1 (877) 438-5566
info@fullonconsulting.com
Preparing ERP data and processes for AI agents

SAP and NetSuite Just Put AI Agents in Your ERP. Is Your Data Ready to Let Them Work?

SAP and NetSuite just launched AI agents that run real ERP workflows. Here's a practical readiness check on data, process ownership and pricing to do before you turn them on.

By Donald D. Hook — Former CTO & CIO | Founder, Full On Consulting | Central Florida
Published

SAP and NetSuite both spent the last two weeks of October telling customers that AI agents are ready to run real work inside the ERP. Not just answer questions. Run workflows.

At SAP TechEd, SAP showed Joule agents that coordinate multi-step processes across apps and data, with a new AI gateway for governance planned for Q1 2027. At SuiteWorld, NetSuite launched industry-specific SuiteAgents for software, product, services, construction, healthcare and nonprofit organizations, with the first three available now.

The demos are impressive. But in my experience, the demo is never the hard part. The hard part is what the agent finds when it starts reading your actual data.

The Readiness Gap Nobody Puts on the Slide

The vendor pitch assumes clean master data, clear process ownership and documented approval rules. Most mid-market ERPs I've walked into have none of those in good shape.

The research backs this up. Fivetran's 2026 Agentic AI Readiness Index found:

  • •Only 15% of organizations are fully ready to run agentic AI in production.
  • •41% already have agents in production anyway.
  • •42% name data quality and lineage as their biggest obstacle.

An agent that reconciles inventory, flags project overruns or chases grant compliance is only as good as the records underneath it. If your item master has three versions of the same part, the agent will act on all three, quickly and confidently.

Five Questions to Answer Before You Turn Anything On

Before an agent touches a live ERP process, I want clear answers to these:

  • •Which process, exactly? Pick one, with a measurable pain: days to close, unbilled hours, stockouts, late payment applications. “Use AI in finance” is not a process.
  • •Who owns it? A named business owner, not IT, who decides what the agent may do and signs off on results.
  • •Is the data trustworthy? Check duplicates, missing fields and stale records in the master data the process depends on.
  • •What can the agent do without a human? Write down the decision boundaries: read only, recommend, or act. Start with recommend.
  • •What does it cost? Get the licensing, consumption and add-on terms in writing. Neither SAP nor NetSuite has made agent pricing the headline, and that's exactly where 2027 budgets get surprised.

Why This Matters Now, Not Next Year

Three things are landing at the same time this quarter:

  • •2027 budgets are being locked. If agents are in your vendor's renewal pitch, they need to be in your plan, with a real business case.
  • •SAP ECC customers are on the clock. Mainstream maintenance ends in 2027. Every month spent on an agent pilot on top of a system you're about to replace is a month not spent on the migration.
  • •Vendors are switching features on. Some AI capabilities show up in your tenant whether you planned for them or not. Someone needs to decide which ones get used, and how.

What a Good First 90 Days Looks Like

I've led ERP programs and the M&A integrations that test them hardest. The companies that get value from AI in ERP follow the same pattern:

  • •Weeks 1–3: Pick one process. Baseline today's metric. Profile the data it depends on.
  • •Weeks 4–6: Fix the worst data issues. Document approval rules and exceptions.
  • •Weeks 7–10: Run the agent in recommend-only mode alongside your team. Compare its output to theirs.
  • •Weeks 11–13: Decide to expand, adjust or stop, based on the metric, not the demo.

That last step matters. I wrote about knowing when to stop an AI pilot earlier this month. The same discipline applies here. For a platform-by-platform view of what SAP, Oracle, Microsoft, Salesforce and NetSuite offer, see my AI in ERP guide for CIOs.

What to Do Next: A Checklist

  • ☐Ask your ERP vendor or partner which AI agents are available on your current version and license.
  • ☐Get agent pricing in writing before your next renewal conversation.
  • ☐Pick one high-pain process and name its business owner.
  • ☐Run a data quality check on the master data that process depends on.
  • ☐Write down decision boundaries: read, recommend or act.
  • ☐If you’re on SAP ECC, confirm your migration timeline before you fund any agent pilot.
  • ☐Put a 90-day go / adjust / stop decision on the calendar.

Frequently Asked Questions

Are AI agents in ERP ready for mid-market companies?

The technology is arriving quickly. SAP and NetSuite both announced AI agents that run multi-step ERP workflows in October 2026. The bigger question is whether your data and processes are ready. Agents act on the records they find, so duplicate master data, unclear process ownership and undocumented approval rules are the most common reasons agents fail to deliver value.

What should we check before turning on an ERP AI agent?

Answer five questions: which single process the agent will support, who owns that process on the business side, whether the underlying data is trustworthy, what the agent may do without a human (read, recommend or act), and what it will cost in licensing, consumption and add-ons, in writing.

Should SAP ECC customers pilot AI agents before migrating to S/4HANA?

Usually not. SAP ECC mainstream maintenance ends in 2027. Confirm your migration timeline before funding an agent pilot on a system you are about to replace, and use the data cleanup work to benefit both the migration and future AI use.

Let's Talk

Every ERP vendor will tell you their agents are ready. The real question is whether your data, your processes and your people are.

I help mid-market companies and nonprofits across Orlando, Leesburg, The Villages and the rest of Central Florida answer that question honestly, with no vendor to sell. If you want an independent read on where AI in your ERP will pay off, and where it won't yet, learn more about our Enterprise Applications and ERP consulting.

Schedule a Strategy Call

Related Reading

Keep Reading

Copyright © 2026 Full On Consulting
info@fullonconsulting.com
Privacy Policy
 
Free CIO Assessment Tool
Schedule a Free Consultation