AI field service scheduling that plans each day around skills, certifications, SLAs, van stock and live traffic, then re-plans when a job overruns. Dispatchers stay in charge in ServiceNow, Salesforce Field Service, Dynamics 365 or ServiceTitan.
AI field service scheduling is software that assigns and sequences service jobs using optimisation and machine learning instead of manual dispatch boards. It weighs technician skills, certifications, parts on the van, SLA deadlines, job duration predictions and traffic. It outputs a daily schedule and route per technician, predicted arrival windows, parts to pick before leaving, and live re-plans when jobs run long or emergencies arrive.
Manual dispatch optimises for who is free, not who can fix it. Parts, skills and realistic job durations get checked too late, so the technician arrives, diagnoses and books a return visit.
Of service calls need at least one additional visit. Unresolved calls needed 1.6 more dispatches on average, per Aberdeen Group.[1]
Of second visits are caused by missing or incorrect parts. Another 25% trace to the technician lacking the right experience.[1]
Of customer complaints: the technician could not resolve the issue. Waiting for appointments followed at 51%, in Aberdeen's 2016 research.[2]
First-time fix when every call is triaged, versus none. Matching the job to skills and parts before dispatch pays off, per Aberdeen.[1]
Models read the ticket, asset history and fault codes to predict likely cause, skills needed, parts needed and realistic duration.
An optimisation engine builds each technician's day, balancing SLA deadlines, travel, overtime rules and preventive maintenance windows.
When a job overruns, a tech calls in sick or an emergency arrives, the day is re-optimised and sent to the dispatcher for one-click approval.
Read-only access to 12 months of work orders, technician skills, van stock and GPS or timesheet data. We agree the pilot metric and a comparison region.
The optimiser re-plans past weeks using what was known at the time. You see travel, SLA and skill-match results next to what actually happened.
Dispatchers in the pilot region receive daily plans and re-plans inside their FSM tool. They accept, edit or reject; every decision is logged.
First-time fix, SLA compliance, drive time and overtime versus baseline and the comparison region, plus the rollout plan and Run cost.
| Criterion | Manual dispatch board | Built-in FSM scheduler | Stratgik build + run |
|---|---|---|---|
| Skill and certification matching | From the dispatcher's memory | Yes, if skills are maintained | Yes, plus predicted skills from the fault |
| Parts availability | Phone call to the van | Rarely considered | Van and depot stock as a constraint |
| Job duration | Flat estimates | Flat or average by job type | Predicted per job, asset and technician |
| Live re-planning | Manual, reactive | Available in premium tiers | Continuous, dispatcher-approved |
| Best fit | Under 10 technicians | Standard jobs with clean data | Mixed skills, strict SLAs, parts-heavy work |
| Replaces your FSM | Not applicable | It is your FSM | No, it works inside it |
Demand for trades such as HVAC is growing faster than the workforce. Getting more fixed jobs from the same technicians is the cheapest capacity you can add.
Two gains: fewer return visits and less drive time per technician. The uplift and minutes saved are assumptions; the pilot measures both in one region.
Illustrative estimate using your inputs and stated assumptions, not a quote or guarantee. The pilot measures the real figure against your baseline.
One region, schedule replay plus live recommendations over 4–6 weeks
Daily scheduling and live re-planning for up to 75 technicians
Multi-region fleets, subcontractors and preventive maintenance programmes
Routing and traffic API calls, cloud compute and model usage billed at cost with no markup. Taxes excluded. GBP and AED prices are indicative conversions from USD.
AI field service scheduling works by predicting what each job needs, then solving for the best assignment and route. Models estimate job duration, required skills and likely parts from ticket text and asset history. An optimiser builds each technician's day around SLAs, certifications, van stock and traffic, and re-plans during the day when conditions change, with the dispatcher approving changes.
Yes, mainly by sending the right skills and parts. Aberdeen Group found 51% of second visits were caused by missing or wrong parts and 25% by technician experience, both of which scheduling can address. Triaging every call before dispatch was associated with 86% first-time fix versus 62% without triage. The pilot measures your own change in one region.
No. It works inside your existing field service management system. We read work orders, technicians, skills and stock through the platform's APIs and write back assignments and schedule changes. Dispatchers keep using ServiceNow, Salesforce Field Service, Dynamics 365 Field Service, ServiceTitan or SAP FSM; they simply get better recommendations and live re-plans.
A Stratgik AI field service scheduling pilot costs $22,000 for one region over four to six weeks. Running daily scheduling and live re-planning for up to 75 technicians costs $5,000 a month, and multi-region Scale starts at $12,000 a month. Routing API and cloud usage are billed at cost. There are no per-technician licence fees.
We need 12 months of work orders with job types, times and outcomes, a technician list with skills and certifications, working hours and locations. Van stock, parts usage, GPS or timesheet data and SLA contracts make the results much better. Missing skills data is common; we can infer a first version from which technicians have closed which job types.
Adoption is highest when people can see why a plan was made and can override it. Every recommendation shows its reasons, such as certification required, part on van or SLA deadline. Dispatchers approve re-plans, and technician preferences and fairness rules are built in. Overrides are logged and reviewed, which usually reveals constraints the data was missing.
See how your past weeks would have been scheduled with skills, parts and SLAs in the plan, before committing to a live pilot.