Understand people & business
Which behaviour, which decision, which value is really at stake? Context before technology.
Use cases
The tool isn't what makes the difference.The system around it is.
HuBizTech never starts with the technology. It starts with people and business — and only then with the tool. Design thinking here means: get the real problem sharp first, then design the system where AI, IoT and automation reinforce each other. The examples below show how those parts come together into something that works.
From idea to working system
Frame the real problem
Not 'we need tool X', but 'this hurts and this is the outcome we need'. Sharpen the problem, not the solution.
Design the system
How do AI, IoT and automation connect? One chain from data to decision to action.
Integrate & enable adoption
Technology that fits how people work — and that they trust and keep using.
Real examples
A mix of my own work and examples from industry — meant to inspire and to show how the parts work together.
EducationFrom my work
Intake automation for an education organisation
Enrolments went through 6 systems, with a 3-day turnaround and high manual error rate.
View caseIndustryFrom my work
IoT energy monitoring for manufacturing SME
Energy consumption per machine unknown; reports only at invoice level, too late for steering.
View caseHealthcareFrom my work
Knowledge portal with retrieval for a healthcare group
Protocols spread across 3 systems; staff rarely found the right version at the right time.
View caseProfessional servicesFrom my work
AI leadership programme for a management team of 30
Management saw AI as an IT project; risk of wrong delegation and wasted investments.
View caseHoldingFrom my work
ICT consolidation post-merger
Two parallel stacks, duplicate licences, conflicting identity and data models.
View caseFinance / corporateFrom my work
Reporting automation for finance
Monthly close took 9 working days, relied on 3 spreadsheets per entity and manual consolidation.
View caseManufacturing · IoTIndustry example
Sensors that predict failure before the line stops
General Motors fitted assembly-line robots with IoT condition sensors; an AI model flags wear before a breakdown. The result: around 15% less unplanned downtime and roughly $20M saved per year.
The lessonThe value isn't the sensor — it's the loop: sensor → model → maintenance decision.
Provalet — predictive maintenance casesProduction · AI & automationIndustry example
From two days to four hours: AI that drafts the paperwork
An automotive supplier replaced manual specification writing with an AI pipeline. Drafting dropped from ~2 days to ~4 hours; people shifted from typing to reviewing exceptions.
The lessonAI drafts, people decide. The win is in that division of roles.
MindStudio — AI documentation workflowsHealthcare · human-centered AIIndustry example
AI that backs the nurse, not replaces her
Across 166,000+ ER visits, the KATE triage model predicted acuity ~27% more accurately than the average nurse (76% vs 60%) — used as a second opinion that catches under-triage.
The lessonGood AI keeps the human deciding — and makes that decision better.
Study on AI triageHow I think
Short notes on systems, AI and the choices behind them.
Leadership & ChangeComing soon
Don't start with tools. Start with thinking.
The most expensive AI projects are those that started implementation too fast. What you need first is a workable thinking framework.
Appearing soonIoT & Smart SystemsComing soon
Why IoT projects fail on integration, not sensors.
The pilot runs. Data comes in. But six months later nobody looks at the dashboard. Four causes — and what solves them.
Appearing soonAI & AutomationComing soon
From scattered tools to one workable system.
A pattern I keep encountering: 14 SaaS tools, 3 spreadsheets and no overview. How to get out without rebuilding.
Appearing soonProduct & DeliveryComing soon
When a product owner needs to be technical enough.
Not every product owner needs to code. But without understanding architecture, you lead the team into tricks instead of choices.
Appearing soonStay in the loop.
A brief digest with new insights, cases and tools I've actually tested. Not a newsletter with 12 topics — just one sharp idea per month.
No spam. Unsubscribe with one click.
Curious how this works for you?
Let's hold your situation up against this way of thinking.