At Housing 2026 this week, Bee had a conversation that stuck with her. A representative from PTSG, a national contractor with around 2,500 engineers working across fire safety, building access, electrical and roofing, mentioned almost in passing that they'd recently introduced AI to generate progress reports from engineer photos and notes. The engineer takes photos on site, adds some notes, and the system turns it into a client-ready report. No more cleaning up scribbles at the end of the day. No more lag between the job and the paperwork.

It's a small shift in how one company handles one type of document. But it's a signal worth paying attention to.

The report problem in housing

Anyone who's worked with field-based teams knows the pattern. The operative or surveyor does the difficult, skilled work on site. Then they come back and spend a disproportionate amount of time writing it up. Notes get reconstructed from memory. Photos get matched to observations made hours earlier. The report that needs to go to the client, or the housing association, or the compliance record ends up being a second job.

For surveyors in social housing, this problem is particularly acute. A stock condition survey, a damp and mould inspection, a void assessment, a HHSRS check: each one generates documentation that needs to be accurate, complete, and defensible. The Regulator of Social Housing is paying close attention to how landlords evidence their compliance with the Consumer Standards. A report that's incomplete, inconsistent, or filed late isn't just an inconvenience. It's a liability.

What changes when AI writes the first draft

The shift PTSG described isn't about replacing the surveyor's judgement. It's about reducing the friction between what happens on site and what gets recorded. When AI can take structured inputs, photos, voice notes, a checklist, and produce a coherent draft report in real time, surveyors spend less time on administration and more time on the inspection itself.

That's the practical case. But there's a compliance case too. A consistent report structure means nothing gets missed. A prompt at the right moment means the surveyor captures the details that matter for HHSRS scoring or Awaab's Law compliance, not just the ones that happened to feel important in the moment. And a report that's generated on site, rather than reconstructed later, is more accurate.

How Alix approaches this

Alix's Walk-and-Talk tool is built around exactly this idea. Surveyors capture observations in the field, with voice, photos, and structured prompts guiding them through the inspection. Alix generates the report automatically, pulling together everything captured on site into a structured output ready for review and sign-off. The surveyor stays focused on the inspection. The documentation takes care of itself.

It's not magic. It's just removing a step that was always unnecessary: the step where skilled people had to translate their own work into a format the system could use.

The direction of travel

What was striking at Housing 2026 was how many conversations came back to the same theme. AI isn't replacing the people doing the work. It's handling the documentation, the structuring, the follow-up, the record-keeping, so that the people doing the work can do more of it. For housing associations under pressure to evidence compliance, manage backlogs, and demonstrate value, that's not a nice-to-have. It's the direction of travel.

If you'd like to see how Alix supports surveyors in the field, we'd be happy to show you.