Workflow redesign, process mining control layers, and resilient network design define agentic operations

By DripPublished

The gist

Operations teams are shifting from automating tasks to redesigning workflows, instrumenting processes, and hardening networks against disruption.

This week’s developments

Workflow Redesign Is the Gatekeeper for Agentic Operations

project44 and Aera, plus product moves from Google, Tulip, and ERP vendors, showed that agentic operations now depend on redesigning workflows before deploying AI. project44 said its predictive intelligence cut freight costs 4%, improved billing and documentation accuracy 22%, reduced check calls 60%, sped exception response 72%, and lowered detention and demurrage fees by up to 99% for one customer. Aera introduced a closed-loop decision automation platform that combines prediction, recommendation, execution, and learning in one system. Google’s Gemini Agent, Tulip’s AI suite for factory operations, and native ERP orchestration all point in the same direction: agents are being embedded inside core enterprise and plant systems, not bolted on as chat interfaces.

The practical shift is from AI that supports work to governed workflows run through AI-orchestrated systems. Vendors are packaging orchestration, automation, and governance together so agents can act only in bounded steps, while deterministic logic handles validation, routing, and approvals. For operations teams, the competitive edge is no longer just adopting an agent; it is redesigning controls, handoffs, and exception paths so those agents can operate safely at scale. That makes process mapping, control design, and human-in-the-loop governance core career skills, not back-office cleanup.

How should we redesign workflows before deploying agentic AI?

If you're an individual contributor

  • AI won’t replace your workflow knowledge; it will expose it.
  • Learn process mapping and exception handling now—your value shifts to supervising AI, not doing the routine steps.

Sources

If you manage a team

  • Your team’s edge is moving from throughput to workflow control.
  • Coach people on handoffs, approvals, and AI oversight; the team that can redesign exceptions will outperform the one that just uses tools.

Sources

If you lead the organization

  • Agentic ops fails if your workflows stay manual and fragmented.
  • Invest in orchestration, governance, and process redesign before more AI spend; talent and operating model need to change together.

Sources

Process Mining Becomes the Control Layer for Agentic Operations

Infor said this week it is expanding agentic AI across manufacturing, distribution, service, projects, supply chains, workforce, finance, and customer service. The most concrete signal is the link to Process Mining & Operational Intelligence: event-log data is being used to spot bottlenecks, prioritize interventions, and trigger multi-step actions with human oversight.

Infor is also placing the release inside Infor OS and the Infor Industry Cloud Platform, with an Agentic Orchestrator coordinating agents across systems, data, and models. That makes this more than generic automation. Infor is tying process intelligence, ERP, automation, and data fabric into one stack so exceptions can be handled where work happens instead of being pushed into separate tools.

For operations teams, the practical shift is clear: process-mining insight is moving from analysis to execution across supply chain, finance, service, and manufacturing. If you run operations, this is a cue to rethink where bottlenecks are resolved, who owns intervention, and how much manual coordination your team still carries between detection and action.

How should we redesign process mining to trigger action, not reports?

If you're an individual contributor

  • Process mining is moving your work from spotting issues to fixing them.
  • Build skill in exception handling and AI oversight; the value now is acting on signals, not just analyzing them.

Sources

If you manage a team

  • Your team’s edge shifts from reporting bottlenecks to resolving them fast.
  • Coach people on intervention playbooks and human-in-the-loop judgment, or they’ll stay stuck in analysis mode.

Sources

If you lead the organization

  • Manual coordination is becoming a design flaw, not an operating norm.
  • Rework ownership, tooling, and talent around execution in-flow; invest where process intelligence can trigger action.

Sources

Network Design Is Shifting from Cost Efficiency to Failure Tolerance

The reported Yandex data-center drone attack is a concrete reminder that even advance preparation cannot fully prevent service disruption when operations depend on concentrated infrastructure. BBC Russian said Yandex had staged equipment ahead of time, yet still suffered interruptions after an earlier attack on 8 October. The lesson is blunt: physical shocks can break continuity faster than contingency plans can absorb them.

That same shift shows up in this week’s research. AJG says companies are adding contingencies, redundancy, and extra stock despite higher operating costs. BCG reports more multiple-sourcing models, with intermediaries, joint ventures, and contract manufacturers used to reduce dependence on a single chain. EY points to segmented supply networks and asset relocation in response to tariffs and geopolitical risk. The pattern is clear: resilience and optionality are increasingly outranking pure cost efficiency in network design.

For operations leaders, this changes how your work is judged. Network decisions now need to prove failure tolerance and recovery speed, not just unit cost. That makes scenario modeling, supplier diversification, rerouting playbooks, and tighter coordination across procurement, logistics, and IT more valuable, especially in semiconductors, automotive, electronics, pharmaceuticals, and logistics.

How should we redesign networks for failure tolerance, not just cost?

If you're an individual contributor

  • Cost-cutting ops skills are losing value; resilience work is the new edge.
  • Get good at scenario checks, reroute planning, and supplier risk review—those are the tasks that make you harder to replace.

Sources

If you manage a team

  • Your team is being judged on recovery speed, not just smooth execution.
  • Coach people on exceptions, contingency playbooks, and cross-functional coordination; that capability now matters more than perfect process.

Sources

If you lead the organization

  • Network design now needs to prove failure tolerance, not just low unit cost.
  • Rebalance investment toward redundancy, multi-sourcing, and segmented networks; your operating model must absorb shocks, not just save money.

Sources

Stay ahead in Operations

Get the weekly Operations brief in your inbox — the developments, what they mean by seniority, and what to do next.