YouTube
Practical walkthroughs
Short demonstrations of EAM, interface behaviour, operational automation and realistic AI/ML use cases.
Open YouTubeField notes & selected topics
Field notes on EAM, interfaces and applied AI
Walkthroughs and selected topics across three independent practices: EAM last-mile engineering, interface reliability and applied AI/ML designed for day-to-day use.
Where I publish
Videos, posts and longer articles
I publish on YouTube, LinkedIn and the external blog. This page brings the main themes together and links each one to the relevant service.
Pilot & validation
Proof belongs close to the operational workflow
A useful pilot should show what works, what fails and which data or process gaps need to be fixed before automation is hardened.
LinkedIn
Field notes and takeaways
Operational observations from EAM, interfaces, data quality, applied AI and partner delivery work.
Open LinkedIn Blog
From pilot to production
Longer notes on EAM last-mile work, interface reliability, data readiness, applied AI/ML and practical automation.
Open BlogFrom the blog
Operational thinking in full articles
English-language articles on ownership, interfaces, data quality, work platforms and the consequences that appear after a solution reaches production.
EAM operations
↗Closing a Work Order Is a Data Quality Decision
5 min read English
Closing a work order establishes the maintenance history that planning, reliability, costing and future decisions will treat as evidence.
- EAM
- Maintenance
- Operations
Reporting
↗Why Reports Become Shadow Systems
5 min read English
Reports are useful for visibility. They become risky when they start replacing process ownership, exception handling and operational control.
- Enterprise Systems
- Operations
- Reporting
Data quality
↗Why Master Data Is Not a Cleanup Project
5 min read English
Master data is not a one-time cleanup project. In production, it is an operating responsibility that shapes workflows, interfaces and reports.
- EAM
- Operations
- Ownership
Work platforms
↗Why Workflow Changes Are Never Local
5 min read English
A workflow change may be made in one screen, but it affects roles, interfaces, reports, open records and recovery paths across the platform.
- Automation
- EAM
- Enterprise Systems
Operations
↗Why Exceptions Are Not Edge Cases in Production
5 min read English
In production, exceptions are not rare edge cases. Reliable automation needs visible exception states, clear ownership and a defined recovery path.
- Operations
- Production
- Reliability
Interface reliability
↗Why Monitoring Is Not the Same as Reconciliation
5 min read English
Monitoring shows whether a process ran. Reconciliation shows whether the business state is still correct. Reliable systems need both.
- Integration
- Operations
- Reliability
Architecture
↗Why Compatibility Is Not the Same as Replaceability
4 min read English
Compatibility can reduce migration effort, but it is not replaceability. Systems become hard to substitute when hidden behaviour accumulates around them.
- Architecture
- Enterprise Systems
- Integration
Interface engineering
↗The Right Structure for an Unreliable Interface
5 min read English
How to choose between stabilising the existing interface, introducing middleware and building a deliberately narrow custom integration.
- Interfaces
- Middleware
- Reliability
Automation
↗Why Automation Fails at the Interfaces, Not the Logic
5 min read English
Why automation often fails at interfaces—not logic—and what teams should check before go-live in real production environments.
- Automation
- Interfaces
- Production
Production readiness
↗The Minimal Guardrails for AI and Automation in Production
5 min read English
A minimal set of guardrails for moving AI and automation from prototype to production without turning quick wins into maintenance debt.
- AI
- Automation
- IT Governance
Maintenance
↗Why ‘Just a Script’ Becomes Long-Term Maintenance
5 min read English
Why a small script often becomes long-term maintenance debt—and how to decide early when automation needs ownership and a lifecycle.
- Automation
- Maintenance
- Ownership
AI ownership
↗Why AI Needs Ownership More Than Intelligence
4 min read English
Why AI initiatives fail not because of intelligence, but because no one clearly owns what happens once the system reaches production.
- AI
- IT Governance
- Ownership
Applied AI
↗ChatGPT Is the New ‘Doctor Google’ – Why DIY AI Fails in Real Projects
4 min read English
Why AI-generated solutions often work in demos but fail in real projects—and how to use AI without creating future technical debt.
- AI
- Automation
- ChatGPT
Topics to explore
Topics from my work
These are service-linked topics, not articles. Use them to understand the problem area, then follow the relevant service path if it matches your situation.
EAM topic
The last mile is where EAM projects usually get expensive
Why work order logic, reports, status flows, interfaces and user workflows often determine whether an EAM setup works in day-to-day operations.
Explore topicInterface topic
A successful API call can still be an operational failure
Operational reliability depends on data meaning, timing, validation, retries, ownership and downstream results—not only an HTTP success response.
Explore topicAI readiness topic
Before an AI pilot, define the decision it should improve
A useful AI/ML use case starts with a real workflow, a baseline, representative data and explicit acceptance and stop criteria.
Explore topicApplied ML prototype
Water-pressure monitoring: separating normal pump cycles from small leak patterns
In a self-initiated prototype, a pressure sensor, Arduino and Linux mini PC captured the time series; ML separated normal water-use and pump-restart cycles from persistent patterns consistent with small leaks.
Explore topicApplied ML prototype
Hoof image screening: flagging visible anomalies for human review
A self-initiated computer-vision prototype used a camera in a barn passage to capture hooves as horses walked past and flag visible irregularities for human review—not to make a veterinary diagnosis.
Explore topicWorkflow topic
Sometimes the useful tool is not another EAM module
Small operational apps can help with requests, checklists, handovers or reporting when the core system should remain stable but day-to-day workflows still need improvement.
Explore topicHave a specific EAM, interface or AI/ML question?
Send the real example. The work is most useful when it is tied to a work order, interface record, data sample, document, decision or workflow.