AI
AI for remote inspections and virtual support
AI runs through the whole session in Virtual Inspection Pro — planning what to capture, reading the live video, extracting the text on the label, and drafting the report at the end. All of it grounded in your own documentation, all of it metered so you can see what it costs.
On the call
While the session is live, AI works next to the agent instead of after the fact.
AI Assist
Reads the live video, names the equipment and the symptom, and pulls the matching article out of your knowledge base with a citation and a match score.
AI Assist in depth →In-call OCR
Freeze a frame and lift serials, model numbers, meter readings, and error codes straight off the label — then copy, share, or save them to the record.
OCR extraction →Voice notes
Dictate findings while your hands are busy. Speech is transcribed into session notes attached to the right photo.
Photo analysis & risk scoring
Captured media is analyzed for defects and integrity signals, feeding the risk score reviewers see on the inspection.
Fraud & tamper detection →Before the call
Set the session up so the right evidence gets captured the first time.
AI template builder
Describe the inspection in plain language and get a structured template — pages, sections, and questions — ready to edit and reorder.
AI walkthroughs
Generates a guided capture plan for the job, then maps each capture back to the checklist item it satisfies.
After the call
The written output writes most of itself.
AI-drafted reports
Executive summaries and prioritized findings drafted from the actual evidence, dropped into your report layout for review.
Automated reports →AI insights
Patterns across sessions — recurring faults, repeat sites, unresolved issues — surfaced where reviewers work.
Running it responsibly
AI you can hand to a whole team without losing control of spend or sourcing.
Usage dashboard & budgets
Per-workspace AI budgets, alerts, and a usage breakdown by feature so cost never arrives as a surprise.
Grounded answers only
Retrieval runs against your own indexed documentation, and every suggestion shows the source it came from.
Cost guardrails
Change-triggered frame analysis, pause when the tab is hidden, and per-session caps an agent can override deliberately.
Common questions
- Where do the AI answers come from?
- From your own manuals, articles, and troubleshooting content. Documentation is indexed for meaning, and every suggestion cites its source with a match score.
- Does the customer need an app?
- No. Customers join from a secure link in their mobile browser. All AI processing happens on your side of the session.
- How do we control AI cost?
- Each workspace has an AI budget and per-session caps. Live analysis only runs when the scene changes, and the usage dashboard breaks spend down by feature.
- What happens to the media AI looks at?
- Frames used only for live analysis are transient. Captured photos and videos follow your existing evidence, retention, and integrity rules.
- How accurate is it?
- Every read carries a confidence score. Low-confidence results are labelled as such and paired with follow-up questions rather than presented as conclusions.
- Which plans include AI?
- AI features are entitlement-gated by plan, with budgets and alerts per workspace. Pricing shows what is included at each tier.