Skip to main content

Submitting a Job for AI Dispatch

Updated today

How Jobs Trigger AI Dispatch

When you create a job in FieldCamp, you have two ways to get AI-powered scheduling recommendations: automatic dispatch when the job is created, or manual re-optimization later when you need to adjust schedules. Both methods use the same intelligent routing engine that analyzes your team's skills, location, availability, and current workload to find the best technician match.

This guide explains how both flows work, what information the AI needs, and what happens during processing.


Automatic Dispatch: Triggering AI on Job Creation

When AI Dispatcher is enabled in your workspace settings, every new unassigned job automatically triggers an optimization request. The moment you click "Create Job," FieldCamp sends a webhook to the AI Dispatcher service containing:

  • The new job details — customer location, required skills, preferred time window, estimated duration

  • Your current schedule — all confirmed technician assignments for that day

  • Pending suggestions — any AI recommendations waiting for your approval

  • Technician availability — who's working, their schedules, skills, and home locations

The AI processes this data in 2-5 seconds and returns a suggestion showing which technician should handle the job and when.

Why include pending suggestions? When multiple jobs arrive within seconds of each other, pending suggestions prevent double-booking. If Job A suggests Bob at 10 AM and Job B arrives before you accept Job A, the AI sees Bob's 10 AM slot as "occupied" and won't suggest him twice.

What Triggers Automatic Dispatch

Automatic dispatch fires when:

  • You manually create a job without assigning a technician

  • A customer books online through your FieldCamp booking page

  • A job is imported from another system (API, CSV upload, integration)

  • An email-to-job conversion creates an unassigned job

The trigger does not fire if you create a job and immediately assign it to a specific technician — the AI assumes manual assignment was intentional.

Job Information Required

For the AI to generate accurate suggestions, each job must include:

Required Field

Why It Matters

Example

Customer location

Calculates travel time from technicians

123 Main St, New York, NY 10001

Required skills

Matches only qualified technicians

HVAC Certified, EPA 608

Preferred time window

Schedules within customer expectations

Tomorrow, 9 AM - 12 PM

Estimated duration

Prevents schedule overlaps

2 hours

Priority level

Weighs urgency vs efficiency

Standard, Urgent, Emergency

Optional but helpful fields:

  • Preferred technician — customer requests specific person

  • Job type — Installation, Repair, Maintenance, Inspection

  • Required equipment — ladder truck, specialty tools

  • Customer notes — special access instructions, pet warnings

Missing location? If a job lacks a valid address, the AI cannot calculate travel times and will leave it unassigned with a "location required" flag. Always verify addresses are complete with city, state, and zip code.

What Happens During Processing

Behind the scenes, the AI Dispatcher:

  1. Validates data — Checks that required fields are present and location is mappable

  2. Identifies candidates — Filters technicians who have required skills and work on the requested day

  3. Checks availability — Removes technicians with schedule conflicts or at capacity limits

  4. Calculates travel times — Measures actual driving distance from each technician's current location

  5. Scores each option — Weighs skills match (35%), proximity (25%), availability (20%), workload (10%), route efficiency (10%)

  6. Generates suggestion — Returns top recommendation with confidence score and reasoning

  7. Stores pending — Adds suggestion to your review queue until you approve or reject

Processing typically completes in under 5 seconds. You'll see the suggestion appear in your AI Dispatcher dashboard with a blue "AI Suggestion" badge.


Manual Re-Optimization: Adjusting Schedules

Sometimes you need to rethink assignments after initial scheduling:

  • A technician calls in sick and you need to reassign their jobs

  • An emergency job arrives and you want to rebalance the day's routes

  • You rejected an AI suggestion and want to see fresh alternatives

  • Workload became uneven and you want to redistribute jobs

Instead of manually dragging jobs around the calendar, trigger re-optimization to get updated suggestions based on current conditions.

How to Trigger Re-Optimization

  1. Go to AI Dispatcher in your FieldCamp sidebar

  2. Select one or more unassigned jobs (or jobs you want to reassign)

  3. Click "Get AI Suggestions" or "Re-Optimize Schedule"

  4. The AI analyzes current schedule state and returns fresh recommendations

Re-optimization uses the exact same engine as automatic dispatch, but pulls the latest schedule data — accounting for any changes you made since the job was first created.

Batch re-optimization: You can select up to 20 jobs and re-optimize them together. The AI will balance assignments across the group instead of optimizing each job independently, often resulting in better overall route efficiency.

When to Use Re-Optimization

Good use cases:

  • Technician availability changed (sick day, early finish, late start)

  • Customer time window shifted

  • You want to compare alternatives before committing

  • Initial suggestion had low confidence (<70%) and you want fresh options

Not needed when:

  • You're happy with the initial suggestion (just accept it)

  • You already know who should take the job (assign manually)

  • The job was correctly assigned and nothing changed


Understanding AI Suggestions

Every suggestion includes:

Header Information

  • Job number and customer name

  • AI Suggestion badge (blue) — distinguishes from manual assignments

  • Status — "Pending Confirmation" (orange) or "Accepted" (green)

Time Details

  • 🤖 AI Scheduled Time — when the AI recommends scheduling based on optimal routing

  • 👤 Customer Preferred Time — the original request window

  • Travel time — estimated drive from previous job

Time deviation warning: If AI scheduled time is 2+ hours outside customer preference, review carefully. Large gaps usually mean the AI detected conflicts — you may need to call the customer for approval or choose an alternative.

Confidence Score

Each suggestion shows a confidence percentage (0-100%) indicating how optimal the assignment is:

  • 90-100% — 🟢 Excellent match, accept with confidence

  • 70-89% — 🔵 Good match with minor trade-offs

  • 50-69% — 🟠 Acceptable but review considerations

  • Below 50% — 🔴 Suboptimal, check alternatives

Click "Why this assignment?" to expand the reasoning panel showing:

  • Key Strengths — green checkmarks for positive factors

  • Considerations — orange warnings for trade-offs or concerns

  • Decision breakdown — scores for skills (35%), proximity (25%), availability (20%), workload (10%), route efficiency (10%)

Real Example: HVAC Repair

Job: AC not cooling, customer requested 2-4 PM today

AI Suggestion:

  • Assigned: Mike (HVAC Certified, EPA 608)

  • Confidence: 87% (Blue — Good)

  • Scheduled: 3:15 PM - 5:15 PM

  • Travel: 12 minutes from previous job

Key Strengths:

  • ✅ Perfect skills match (EPA 608, residential HVAC experience)

  • ✅ Within customer preferred window (3:15 PM starts in 2-4 PM range)

  • ✅ Efficient routing (previous job is 12 min away)

  • ✅ Moderate workload (Mike at 65% capacity today)

Considerations:

  • ⚠️ Starts 1 hour 15 minutes after customer's earliest preference (customer requested 2 PM, AI scheduled 3:15 PM)

Decision: Accept — arrival is within the 2-4 PM window and all other factors are excellent.


Taking Action on Suggestions

After reviewing a suggestion, you have three options:

✅ Accept

Confirms the assignment. The job locks into the technician's schedule, they receive a push notification on their mobile app, and the customer gets an automated confirmation with arrival time.

When to accept:

  • Confidence 85%+ with no critical concerns

  • All decision factors show green/blue status

  • Time deviation from customer preference is acceptable

  • No better alternative available

❌ Reject

Dismisses this specific suggestion. The system prompts for a rejection reason (skills mismatch, time conflict, workload concerns, route inefficiency). Your feedback trains the AI to avoid similar mistakes in future suggestions.

When to reject:

  • Critical conflict detected (overlapping appointments, impossible travel)

  • Skills don't actually match requirements

  • Technician explicitly unavailable (you know they're on vacation but system doesn't)

  • Customer specifically requested different technician

🔄 Reassign

Opens the alternatives panel showing 2-3 other technician options ranked by confidence. Each alternative displays trade-offs compared to the primary suggestion.

When to reassign:

  • Confidence below 70%

  • Alternative has higher confidence or better fits your preferences

  • Primary suggestion has workload concerns but alternative has capacity

  • You want to balance work across multiple technicians


Common Scenarios

Emergency Job: Burst Pipe

Situation: Customer calls with water spraying everywhere, needs help ASAP

Process:

  1. Create job, mark priority as "Emergency"

  2. Set time window to "Next 2 hours"

  3. Required skills: Emergency Plumbing, Water Main Shutoff

  4. AI immediately suggests nearest qualified plumber

  5. Suggestion shows current assignment + travel time: "Carlos finishing service call 8 min away, can arrive in 15 minutes"

  6. Accept → Carlos gets urgent notification, customer gets ETA text

Confidence factors: Proximity weighted highest for emergencies, skills match critical, route efficiency less important.

Technician Sick Day: Redistributing Jobs

Situation: Sarah calls in sick, has 6 jobs scheduled today

Process:

  1. Mark Sarah as unavailable in Team settings

  2. Go to AI Dispatcher, filter for Sarah's jobs

  3. Select all 6 jobs, click "Re-Optimize"

  4. AI analyzes remaining technicians' schedules and generates new suggestions

  5. Review confidence scores — some may be lower due to tighter capacity

  6. Accept high-confidence suggestions, manually handle edge cases

  7. System sends updated notifications to customers with new technician names and ETAs

Customer Reschedules: Time Window Changed

Situation: Customer originally requested 9 AM, now can only do 2 PM

Process:

  1. Edit job, update preferred time window to 2-4 PM

  2. Click "Get AI Suggestions" (re-optimization)

  3. AI may suggest same technician at new time, or different technician if original is now booked

  4. Review new confidence score — route efficiency may change if 2 PM creates backtracking

  5. Accept or explore alternatives


Best Practices

  • Complete job info upfront — accurate location, skills, duration = better suggestions

  • Trust high confidence — 85%+ suggestions are usually optimal, minimal review needed

  • Review considerations — orange warnings highlight trade-offs worth checking

  • Use batch re-optimization — when multiple jobs need reassignment, optimize together for better routes

  • Provide rejection feedback — helps AI learn your preferences and avoid repeated mistakes

  • Keep team info current — update skills, schedules, and availability immediately when changes occur

  • Call customers for time deviations — 2+ hour shifts from preference usually need approval


Remember: AI Dispatcher suggests — you decide. The system handles the computational work of analyzing hundreds of factors per second, but your operational judgment and customer relationships remain essential to great service.

Did this answer your question?