Model first, business logic second
You stand up YOLO + CoreML locally and see what the model really returns — before anyone writes access rules or alerts.
Local AI Vision · Computer Vision
Local AI Vision from scratch — stand up a model, track objects, generate events
This is not a course on “how to build a parking system”. It is a programme where you build a local Computer Vision pipeline: image → model → detection → tracking → zones → business events. Using a camera, people and vehicles (e.g. gym entry, zone, access) you learn building blocks you can reuse for retail, security, logistics or smart building.
Who it is for
Developers, data engineers, solution architects and technical roles who want a local CV pipeline (Mac) that turns a camera feed into concrete business events — without cloud as a prerequisite.
Benefits
Universal blocks, not one parking case
People and cars are the teaching example; the same blocks (detection → tracking → zones → events) transfer to retail, security, logistics or smart building.
Intelligence in the event layer
The model supplies observations. Enter / exit / zone breach is yours — and you get an event file ready for a dashboard or integration.
What you get
- Standing up a local model (YOLO + CoreML on a Mac)
- Understanding the pipeline: detection → tracking → interpretation
- Event generator (enter / exit / zone)
- Ready project folder — swap the example for your own case
Outcomes
- You know how to take a camera, stand up a model locally and turn image into events (enter, exit, zone breach)
- You understand the gap between model observations and business logic in the event layer
- You leave with a project folder ready to swap people/vehicles for your own case
The example uses people and cars, but **the blocks are universal**. First you stand up the model and see what it actually returns — only then do you build business logic. AI does not “know” what is happening: it supplies observations; intelligence lives in the event layer.
Tailor your enquiry (AI Vision)
Tick context — helps us prepare an open cohort or a closed workshop for your case.
Format
Target case (multi-select)
Your choices are attached when you send the enquiry.
Send this configuration with your message — we will respond with next steps.
Pitch in three sentences
We build a local Computer Vision system from scratch. Using a camera, people and vehicles you learn to stand up a model, track objects and generate events (e.g. gym entry / zone breach). After 8 hours you have a pipeline and a project folder — not slides about “AI vision”.
Blocks → class example → what you can do later
| Block | In class | Later reuse |
|---|---|---|
| Detection | person / car | customers, staff, vehicles, parcels |
| Tracking + ID | CAR #15, PERSON #7 | dwell time, object path |
| Zones | parking / road / entrance | VIP, forbidden, queue, entry |
| Events | vehicle_enter / person_enter | gym access, turnover, alerts |
| Data (CSV/JSON) | events.csv | dashboard, report, system integration |
Areas
- AI & transformation