Issues, quality and site changes
Logs photos, location, owner, due date and proof of resolution. AI helps classify; the technician validates.
What it is
Managing issues, quality and site changes logs each open point with its photo, location, owner, due date and proof of resolution, and tracks it to closure. AI helps classify and group; the technician validates. It builds on the management panel SAB already runs in production.
Who it's for
Site managers, quality control and project supervision who keep punch lists, non-conformities and changes scattered across WhatsApp, emails and spreadsheets.
The problem
Issues get lost among messages: nobody knows how many are open, whose they are or which are due. At the end of the job the proof that they were resolved is missing, right when it is needed to get paid or for a claim.
How it works
The process, step by step.
- 01
Open with evidence
The issue is logged with photo, location and description from the phone, on site.
- 02
Assisted classification
AI proposes type, owner and priority from the photo and text; the technician confirms or corrects.
- 03
Tracking
Each point has a state, owner and due date; what is open, overdue and closed is visible.
- 04
Verified closure
It is closed with proof of resolution, and the history remains for payment or claims.
What you get
The delivered result.
A single log of issues, quality and changes, with photo and location.
State, owner and due date of each point, with overdue items visible.
Proof of resolution on closure, stored and dated.
An AI-assisted classification the technician can always correct.
What it needs
- Your issue types (punch item, non-conformity, change) and their states.
- Who is responsible and the deadlines for each type.
- What counts as proof of resolution (photo, document, signature).
A concrete example
A punch item on the north façade: photo, location “level 3, axis B”, owner “painting”, due in 5 days. AI classifies it as “finish” and proposes medium priority; the site manager raises it to high. When resolved it is closed with the photo of the fix, which stays in the job's history.
What a person reviews
Always the technician. AI suggests type and priority to save typing, but rating, assigning and closing an issue is done by a person; the system does not close points on its own.
Possible integrations
Possibilities subject to assessment until verified.
- · SAB's management panel as a base (already in production), adapted to your types.
- · Export of issue reports for client, supervision or the file.
Limits and conditions
- · AI classification is an aid, not a quality verdict: the technician validates.
- · It does not replace inspections or the site's regulatory responsibilities.
- · It reuses real SAB software; adapting it to your quality flows is part of the service.
FAQ
Frequently asked questions
Isn't this already SAB's panel?
The base is: SAB's management panel is in production and is reused to log, track and close points with evidence. The site-specific part —issue, quality and change types, and photo-assisted classification— is adapted and developed on top of that base.
Does the AI decide if something is well executed?
No. It proposes type, owner and priority to save work, but the quality assessment and the closure are done by the technician. The system does not rule on conformity on its own.
Is the evidence useful to get paid or for claims?
That is one of the reasons it exists: each closure stores its dated evidence, so at the end of the job there is the history that is usually missing when a payment must be justified or a claim answered.
How do we start?
With one issue type and one site. We set up a pilot on the existing panel, check that the log is complete and the closure evidence is usable, and scale to more types from there.
Let's start with one process and one site.
The first engagement is small and demonstrable: one process, concrete data and a measured result. Tell us yours.
Request an automation diagnosis