What the AI in WorkmanIQ does.
WorkmanIQ uses AI for six specific jobs: sorting incoming requests, filling in work order fields, ranking assets by risk, answering reporting questions, drafting SOPs, and looking up data through the assistant. Each one takes a task your staff already does and makes it quicker. People review the results, and the AI does not change your records on its own.
Sorting citizen requests
When a request comes in from the public portal, SeeClickFix, or a webhook, WorkmanIQ suggests a division, category, and priority. Requests with a high confidence score become work orders right away. The rest wait for a supervisor.
- Checks for open work orders and assets within 200 meters
- Flags duplicates and repeat problems at the same location
- Shows a confidence score and a short explanation for each result
Suggested fields
As someone types a work order description, WorkmanIQ suggests the division, category, priority, and work type. The work order goes to the right queue without the person having to look up codes.
- Each suggestion can be accepted or changed
- Runs on a lower cost model because it is used so often
- Uses your own divisions, categories, and roles
Risk scores for assets
A weighted scoring model ranks assets by how likely they are to fail, so crews can get to the problem equipment before it breaks down.
- Looks at age against expected lifespan, condition trend, repair frequency, and warranty
- Rates each asset Critical, High, Medium, or Low with a written reason
- Recalculates as work orders are closed
Questions in plain English
Type a question and get a chart, a sortable table, and a short written summary based on your current data.
- "Which assets cost us the most this quarter?" returns a ranked table with the trend
- "Which crews are overloaded?" compares each technician's workload to the average
- "Which work orders are about to miss their SLA?" sorts by hours remaining
SOP drafts
Upload a manufacturer's manual as a PDF and WorkmanIQ drafts a standard operating procedure. It can also start from just the equipment type and model.
- Numbered steps with safety notes and an estimated time for each
- Required parts, total time, and recommended service interval
- A supervisor reviews and edits the draft before it is published
The WorkmanIQ AI Assistant
A chat window available on any screen in the app. You can ask it about your organization's data the same way you would ask a coworker who has the reports open.
- Covers overdue work, costs by division, asset condition, fleet, staff, and PMs
- Keeps a history of past conversations
- Read only and limited to your organization, so it can look things up but cannot change records
What runs these features.
This part is for IT and procurement reviewers. It lists the models and internal engines the features above depend on.
Language models
- Larger models for the assistant, SOP drafts, and reporting questions
- Lower cost models for suggested fields and request triage, which run far more often
- Each request includes your divisions, categories, and roles, so answers use your names for things
- Your own API key, stored encrypted, is used when you provide one
- Usage limits per user, with a log of each request
Internal scoring engines
- A weighted scoring engine for asset risk
- A location matching engine for triage and duplicate detection
- Both run inside the platform with no outside calls, and give the same result for the same data
See the AI features in a real workflow.
The How It Works page follows a pothole report and a lift station failure from start to finish, including where triage, risk scores, and SOP drafts come in.
See how it works arrow_forward