01 / intake_task
Claim Intake Agent
Data collection
Collects claim type, insured person, event details, driver, third parties, appraiser, garage and required licenses.
Case Study / Insurance Agency — AutoClaim
5 stages, 4 external parties, automated reminders — and one rule: every unclear case goes back to a human.
Built for an insurance agency that handles vehicle insurance claims. AutoClaim automates the operational flow from first accident report to final claim closure: intake, documents, signatures, insurer emails, appraiser and garage follow-up, reminders, dashboard tracking, and human escalation.
autoclaim.flow · live
● running Claim reported └─ intake created ├─ collect required fields ├─ request licenses + bank approval ├─ send signature links └─ monitor completion AI email classifier ├─ clear insurer request → act └─ ambiguous request → human handoff Policy: automation handles the predictable; people handle judgment.
5Workflow stages
4External parties
7dFollow-up cycle
HITLHuman-in-the-loop
The manual baseline
The bottleneck was not one task. It was the constant switching.
Before AutoClaim, every claim required the office team to manually collect client details, chase missing documents, send forms, follow up with the insurance company, coordinate with the appraiser and garage, and remember when to escalate.
The agency worked across WhatsApp, SMS, email, forms, documents, insurers, garages, appraisers, and internal follow-up lists. One missing signature or one unanswered insurer email could quietly delay the entire claim.
AutoClaim converts this fragmented process into a tracked operational pipeline where every claim has a stage, a next action, an owner, and an escalation rule.
The system
One AI workflow that manages the claim from opening to closure.
The system follows the same five-stage operating logic used by the agency: opening and intake, customer follow-up, insurer coordination, appraiser and garage coordination, and final closure.
1
Opening & intakeOffice opens the claim; system collects claim type, insured person, event details, driver, third parties, appraiser, garage and required documents.
2
Customer follow-upDigital signature links are sent by WhatsApp or SMS. Automated reminders continue until documents and consents are complete.
3
Insurance company coordinationThe claim package is sent to the insurer. Incoming emails are monitored, classified and either processed or escalated.
4
Appraiser & garage coordinationThe system requests the appraiser report, appraiser invoice, garage invoice and repair status, then sends reminders when needed.
5
ClosureWhen payment approval is received, the system summarizes the claim and closes it in the control dashboard.
Workflow map: from accident report to final payment approval and dashboard closure.
The claim crew
Not a chatbot. An operations agent with a dashboard.
AutoClaim is organized as a set of operational responsibilities. Each agent handles a narrow part of the claim process, while the dashboard gives management full visibility.
02 / signature_task
Document & Signature Agent
Completion control
Sends signature links to owner and driver by WhatsApp or SMS and tracks whether both consents were signed.
03 / reminder_task
Customer Follow-Up Agent
No silent delays
Sends automatic reminders on Day 1 and Day 2, then alerts the office if documents are still missing.
04 / insurer_task
Insurer Coordination Agent
Email operations
Sends claim materials to the insurance company, monitors replies and forwards clear requests to the client.
05 / classifier_task
Email Intent Classifier
AI triage
Analyzes incoming emails. Clear intent is processed automatically; ambiguous intent is escalated to the office.
06 / garage_task
Appraiser & Garage Agent
External follow-up
Requests appraiser report, appraiser invoice, garage invoice and repair status with timed reminders.
07 / closure_task
Closure Agent
Case completion
Receives payment approval, sends a structured summary to the office and closes the claim in the dashboard.
The control layer
Built for automation, but not blind automation.
The system automates predictable steps and escalates anything that may require judgment. Scenario mapping is a development prerequisite: each party, possible reply, automated response and human-handoff condition must be defined before launch.
handoff_rules.json
{
"missing_documents": "alert_office_after_failed_customer_followup",
"unclear_email": "escalate_to_human",
"non_standard_request": "escalate_to_human",
"decision_required": "notify_dashboard_and_email",
"external_party_delay": "send_reminder_after_waiting_period"
}
claim_package.txt
Required package: · claim notification form · owner + driver consents · vehicle license · driver license · bank approval / check copy · appraiser report · appraiser invoice · garage invoice · payment approval
The dashboard
Every claim has a visible status, stage and next action.
The office no longer needs to ask, “What is happening with this claim?” The dashboard answers it with live status, stage, documents, alerts and communication history.
Open claims tableClaim number, customer, type, current stage, status, insurer, age and required action.
Document checklistVehicle license, driver license, bank approval, signatures and claim forms.
Reminder timelineCustomer, insurer, appraiser and garage follow-up events in one place.
Decision alertsUrgent cases that require human review or non-standard handling.
KPI stripAutomated handling rate, average closure time and cases requiring decision.
Dashboard mockup: claims list, stage indicators, status tags, reminders and urgent alerts.
The result
Faster follow-up, fewer forgotten tasks, cleaner handoffs.
AutoClaim reduces the administrative load by moving repetitive follow-up work from people to an AI-controlled workflow. Human attention is focused where it matters: exceptions, judgment and relationship management.
Less manual chasingAutomatic reminders reduce repetitive calls, WhatsApp messages and email follow-ups.
Fewer forgotten tasksEvery claim has a stage, status and next action instead of relying on memory or scattered notes.
Cleaner document collectionRequired documents and signatures are tracked until the package is complete.
Better management visibilityThe agency owner can see what is stuck, why it is stuck and who needs to act.
Clear audit trailCustomer, insurer, appraiser and garage communication is recorded as part of the claim history.
Human escalation by designAmbiguous messages and decision points are routed to the office instead of being guessed by the AI.
The implementation
Small enough to pilot. Structured enough to scale.
The claim-handling module can start as a focused pilot and then expand into a broader insurance agency automation layer.
#
Stage
Hours
Timeline
1.0
Discovery & requirements
5
Week 1
2.0
Technical design
5
Week 1
3.0
Development
100
Weeks 2–5
4.0
Internal testing & bug fixing
20
Week 6
5.0
UAT and user changes
20
Week 7
6.0
Training and handover
10
Week 8
7.0
Close support, first 30 days
20
Week 9
+
Safety buffer
27
—
=
Total project roadmap
207
9–10 weeks
Recommended pilot: vehicle claims
Start with intake + reminders
Add insurer email classification
Launch dashboard
Improve edge cases during support window
The pattern
Insurance claims are the case study. The architecture is broader.
The same architecture can be reused anywhere an insurance agency has repetitive, document-heavy and follow-up-heavy workflows.
Vehicle claims
Business insurance quote requests
Policy renewal workflows
Customer onboarding
Missing document follow-up
Lead warming through WhatsApp
The domain changes. The pattern stays the same: structured intake, document collection, automated communication, AI classification, dashboard visibility, human escalation and audit trail.
Next step
Want an AI operations layer for your insurance agency?
In a 15-minute fit call, Profitec AI maps your claim workflow, identifies the repetitive tasks, defines the escalation rules and shows where an AI agent can start saving time immediately.
Book a fit call Describe your workflow