We shipped four updates to Inspection Assistant based on what customers told us was missing. Inspection Assistant is the AI review tool built into Inspect Point that checks inspection reports for unanswered questions, missing photos, missing signatures, grammar issues, shorthand, and historical deficiency mismatches before they reach the office or a customer. No separate app, no extra logins.
Here’s what changed and why it matters.
When we launched Inspection Assistant, our mission was straightforward: give fire and life safety technicians an AI co-pilot that catches mistakes before they leave the job site. Fewer callbacks. Cleaner reports. Less admin back-and-forth.
That core promise hasn’t changed, but we know field capture is only part of the story. The office has to review. It should create coaching opportunities. And, ultimately, we want to know where – and how much – this is primarily improving our customers’ day-to-day operations.
1. Insights Dashboard
Before this update, inspection data lived in individual reports. If you wanted to know how your team was performing across the board, whether techs were acting on AI feedback, or which issue types kept coming up, you had to piece it together manually. That rarely happened.
The Inspection Assistant Insights Dashboard changes that. Accessible within our platform’s dashboard, it pulls AI performance data from across your entire operation and surfaces it in two tabs: Overview and Tech Performance.
Overview: Where is AI Providing Value?
The headline chart tracks Issues Caught vs. Resolved over time. Blue is total issues flagged by AI, green is issues corrected, red is issues that escaped without resolution. This chart, watched over weeks, tells you the most important thing: is your team acting on AI feedback, or ignoring it?
Five key metric cards fill out the picture:
- Total Existing Issues: A running cumulative count of everything IA has ever flagged
- Total Open Issues: Issues still waiting for action – if this number climbs and doesn’t come down, the office isn’t reviewing
- Average Processing Time: How long IA takes per inspection
- Average Resubmissions: How often inspections are going back and forth for correction
- Escaped Issues: Issues that made it to the final report unresolved
Escaped Issues is the most important number for proving ROI. A high escaped count means issues are slipping through to customers and AHJs before anyone caught them. A declining “escaped” count is proof the system is working.
Additionally, the Resolution Rate metric tracks the percentage of AI-flagged issues that actually get fixed. In practice, this is a strong indicator of how much value your team is getting from Inspection Assistant.
Two additional charts round out the tab – A breakdown of the top issue categories being flagged (useful for identifying recurring training gaps), and a Dismissed vs. Resolved by Issue Type view. If a specific issue type is being dismissed more than 70% of the time, it’s likely a false positive and worth investigating whether to turn it off. If an issue type is consistently resolved, it’s clearly earning its keep.
Tech Performance: Coaching Built on Data
The second tab shifts from the operation to the individual. Every technician gets a row showing issues flagged, resolution rate, most common issue type, average resubmissions, and a coaching priority status. When a tech’s metrics cross certain thresholds, like a low resolution rate or a high issue count, they’re flagged for attention.
Clicking any technician opens a detailed breakdown of their performance relative to the team average, their top issue types, and specific focus areas. For inspection managers, this is the difference between knowing something is wrong and being able to show exactly what needs to change.
Instead of waiting for a bad inspection to surface a problem, managers can now spot patterns early and coach before the customer notices.
2. Workflow Optionality
Not every team wants to roll out AI review the same way. Some techs want to see feedback directly. Some want the office to handle it. We heard this repeatedly, so we built three new workflows to help you align Inspection Assistant with your team’s process, configurable at the company level.
Tech-Required (default)
Technicians must send the inspection through Inspection Assistant before it goes to the office. Those results are visible to both the tech and the office. Inspections can’t move forward without going through Inspection Assistant review first, which creates the most technician accountability and ensures no report reaches the office without being checked.
Tech-Optional
Technicians can choose to send reports through Inspection Assistant or skip straight to submission. If they skipt, Inspection Assistant still runs automatically, but only the office sees the output. You get full coverage without forcing a new step on every tech.
Office-First
Inspection Assistant runs automatically on every submission, but results only go to the office. The technician never sees that feedback directly; they just receive a normal correction notice if the office decides to send one back. This is the most controlled QA model: technicians experience the correction workflow they already know, without any change to their daily routine, while the office gets a full AI-powered review of every inspection before anything reaches a customer or an AHJ.
Workflows can be set as a company-wide default and then overridden per technician. This way, say a newer hire might run Tech Required while a veteran runs Tech Optional. Or maybe a team running a pilot of Inspection Assistant might want to run Office-First for most technicians, while a few experienced techs try Tech Required.
The right workflow isn’t the same for every team, and it probably isn’t the same for every technician on your team. Now you don’t have to choose one size for everyone.
3. Technician Scorecard
Coaching technicians in fire protection is usually reactive. A report comes back with errors, a customer pushes back, or an AHJ flags something. By then, the pattern has often been repeating for weeks.
The Tech Scorecard gives inspection managers a data-backed view of individual technician performance, pulled directly from Inspection Assistant output. For each technician, it surfaces:
- Total issues flagged by AI on their inspections
- Their personal resolution rate
- Their most common issue type
- Average resubmissions
- A coaching priority status, including “Immediate Intervention” flags for the most critical cases
Rather than subjective impressions, every coaching conversation can start from a shared set of numbers, like which issue types are driving resubmissions, how trends have changed over the past month, and what a healthy resolution rate looks like on the same team. The technician sees the same data, which makes feedback easier to receive and easier to act on.
This is particularly powerful in Tech First mode, where resolution rate is a direct measure of how actively a technician is engaging with Inspection Assistant’s suggestions. A tech with a 10% resolution rate and a 90% dismiss rate is a very different coaching conversation than a tech with a 70% resolution rate who’s struggling with one specific issue category. The scorecard makes that distinction visible.
It may very well be that inspection managers already knew which technicians needed coaching. But now, they’re equipped with the data to show exactly why and the specificity to make the conversation stick.
4. Bulk Actions
Here’s a problem that sounds minor until you’re living it: Inspection Assistant flags issues across thirty inspections submitted overnight. The person reviewing them opens the queue in the morning and has to open each one individually, act on the AI feedback, and route it forward. One by one.
Without bulk actions, that just creates a new bottleneck.
Inside each inspection’s review tab, the Issues table now supports bulk selection. Check the boxes next to multiple issues, then act on all of them at once:
- Suggest Correction: Send selected issues back to the tech in one step
- Dismiss: Close out false positives or low-priority flags without opening each one
- Undo: Reopen previously resolved or dismissed issues in bulk if a decision needs to change
The Issues table’s filters and sorting make bulk actions even more powerful. Filter by Issue Type to isolate a specific category, sort by Impact to surface high-priority issues first, then select and act on everything that needs the same treatment. Your team can clear a full morning’s queue of unanswered questions with a handful of clicks instead of working through them one at a time.
One note worth keeping in mind: bulk-dismissing everything in a queue is a red flag. If your team is bulk-dismissing across the board, it’s worth asking whether Inspection Assistant is flagging the right issues, not whether the bulk action is being used correctly. The goal is to make meaningful review faster, not to make it easier to skip.
The Bigger Picture
These four updates share a common thread – they take the value Inspection Assistant creates in the field and make it visible and actionable for everyone else in the business.
The Insights Dashboard turns inspection data into operational intelligence. On a more granular level, the Tech Scorecard turns AI output into targeted, objective coaching. Workflow controls make sure Inspection Assistant fits your team’s process, not the other way around. And bulk actions let a busy office keep pace with what Inspection Assistant is producing.
What’s Next
Every one of these updates came from customer conversations. Teams told us what was slowing them down, what was blocking adoption, and what would actually make a difference. That’s how we build.
If you’re using Inspection Assistant, we want to hear what’s working and what’s not. If you haven’t tried it yet, request a demo, and we’ll walk through how it fits your specific workflow.
See Inspection Assistant in Action
Get A DemoMore Resources
