Practical AI for real operations

AI in Operations

I connect AI with the decisions people make in production, supply chain, quality, and regulated operations. The work starts with process truth and decision ownership, then moves to models and agents. In that order.

Explore 16 operational decisions See what agents can build

Built on public proof

Digital operations leadership grounded in enterprise execution
Former university lecturer with reusable learning material
19k+ LinkedIn community around practical digital operations
YouTube lectures and GitHub repositories as inspectable proof

Start here

AI in operations starts with decision ownership, not tool enthusiasm.

The useful question is not whether agents can automate something. The useful question is whether the workflow is bounded, the data is reliable, failures are visible, and a human review gate is clear.

Read the AI in Operations section Resource sheet

Use-case readinessBound the workflow, data, failure modes, and human gate

Decision ownershipRemove coordination loops without hiding accountability

OEE before AI hypeProduction AI starts with losses, not models

Evidence library50 public sources for agentic AI in operations

Choose a learning path

AI in Operations

Use agents where workflows are bounded, reviewable, and connected to real operational decisions.

Supply Chain Analytics

Map analytics and AI use cases to planning, sourcing, production, risk, and delivery decisions.

Analytics Translator 2.0

Bridge business, analytics, IT, AI, process, governance, and adoption in the agentic era.

From LinkedIn series to learning path

Sixteen decisions agents should not own alone.

The series follows the operating system: make the product, plan the network, then cross supplier and customer boundaries. Every case names the trade-off and the people who live with it.

Make

Maintenance, vision, scheduling, digital twins, and batch size. Factory physics gets the first and last word.

Explore 5 decisions

Plan

Control towers, forecasts, inventory, segmentation, risk, and network design. Visibility creates sharper trade-offs, not automatic answers.

Explore 7 decisions

Cross boundaries

Supplier trust, compliance evidence, routing exceptions, and customer promises need explicit rights and limits.

Explore 4 decisions

Read the full 16-case journey

Agent workbench

Useful work leaves an artifact a human can inspect.

Not every experiment belongs inside the business story. Some are here because they show what agents can do well: inspect a source, build something bounded, test it, and leave a clean review path.

Open the workbench Try the readiness canvas

Interactive use-case library25+ public HTML presentations across source, make, plan, and deliver

Agent readiness canvasEight practical checks in a browser-only prototype

Hermes Agent field guideTools, skills, scheduled work, and verification

Inspectable sourceRepositories behind the lectures, pages, and prototypes

Resource and proof layer

Inspect the sources, lectures, and working artifacts.

Everything here is currently open: source collections, lecture routes, and GitHub artifacts. Optional email updates and downloadable packages can be added later.

Agentic operations AI resource sheet50 source links for credible agentic AI in operations

Public lecture archiveStructured learning paths across operations, analytics, and AI

supplychainanalyticsSupply-chain analytics use cases

operations_use_case_selectionAI agent readiness workflows

View curated GitHub projects

Talks & professional education

For selected talks, guest lectures, or professional education inquiries around analytics translation, digital operations, agentic AI, physical AI, and regulated operations, reach out directly.

Talks / Contact

Personal educational content by Frank Kienle. Views are personal. Examples are based on public, educational, historical, or synthetic material unless stated otherwise. No employer-confidential, customer-confidential, or supplier-confidential information is shared.