📘 Documentation · Pre-planning for recurring visits

Pre-Planning for Recurring Visits

Turn municipal decisions into approved, ready-to-execute schedules before they reach daily ops. Caire blends OCR ingestion, AI optimization, and customer sign-off to eliminate trial-and-error loops.

50–75% faster onboarding time
80% first-proposal approval rate
75–80% service time utilization
Illustration av Caires AI-hjärtechip för förplanering

Pre-planning of recurring visits

🎯 Overview: Pre-Planning enables home care organizations to optimize recurring visits before they enter the daily scheduling system. Caire replaces manual trial-and-error with AI-driven optimization, saving over 50% time during new customer onboarding while achieving 75-80%+ service time efficiency and 80% first-proposal approval rates. When a plan is approved, the chosen times are written into caregivers’ slingor as locked visits, while new or more flexible visits start out unlocked so that daily optimization can still move them within agreed buffers.

🎨 Interactive UX Mockups

→ Click here to view Caire's interface design for this feature

Mockups show exactly how the pre-planning workflow will look in Caire's liquid glass Apple design. All 8 steps from this document are visualized with a navigable interface.

Pattern Review

Identify recurring visits before client meetings

Recurring visit approval workflow with anonymised client and staff details
Förslagspanelen visar återkommande besöksmönster som AI föreslår att flytta. Varje mönster sammanfattar datum, tider och rekommenderad åtgärd. Alla klient- och medarbetarnamn är anonymiserade för integritet.

Innovation

All tasks start as flexible. Caire adds intelligence by:

  1. PDF upload with OCR: Upload 1 or 300 municipal decisions - AI automatically extracts all data and completely replaces manual entry
  2. Accept all visits with time windows (allowed periods, e.g. "Morning 07:00-10:00" or "Flexible 07:00-22:00" for cleaning)
  3. Find optimal fixed times through AI optimization over weeks/months - biggest impact on weekly/bi-weekly visits (cleaning, groceries, social activities)
  4. Facilitate customer meetings to confirm proposed times
  5. Export finished schedules to operations systems (Welfare/Epsilon or Carefox)
  6. Enable daily optimization with flexibility buffers (±15-30 minutes)

Problems Solved

Traditional "trial loop" (test loop) is replaced by comprehensive AI pre-planning that considers:

Table of Contents

  1. Problem Description
  2. Solution
  3. Core Concepts
  4. Complete User Workflow
  5. System Architecture
  6. Time Windows vs Flexibility
  7. Examples

Problem Description

Current State (Without Pre-Planning)

Home care organizations receive municipal decisions that specify:

Current manual process:

  1. Scheduler/coordinator (or equivalent role) enters visits manually in Welfare/Epsilon with preliminary times
  2. Creates a "trial loop" (temporary test schedule) with a subset of visits
  3. Checks manually if visits fit within staff schedules
  4. Books customer meeting to agree on times
  5. Adjusts schedule manually based on customer feedback
  6. Enters everything again in Welfare/Epsilon/Carefox with fixed times
  7. Daily planning is done manually
  8. Implementation plans are sent to municipality (via municipality's system)

Problems:

  • ❌ Time-consuming: 4-8 hours per new customer onboarding
  • ❌ Suboptimal: "Trial loop" only tests small subsets, misses the whole picture
  • ❌ Manual errors: Re-entering data multiple times
  • ❌ No route optimization: Routes not considered during pre-planning
  • ❌ Reactive: Customer preferences collected without showing feasible alternatives
  • ❌ No prioritization: Mandatory vs optional tasks not distinguished

Solution

Pre-Planning i Caire replaces the "trial loop" with AI-driven optimization that:

  1. Accepts ALL visits with flexible time windows
  2. Optimizes comprehensively considering all staff, skills and trofel
  3. Proposes feasible schedules ready to show customers
  4. Incorporates feedback and re-optimizes when needed
  5. Exports to operations systems for daily execution

Main Principle: Two Types of Flexibility

🔵 PHASE 1: PRE-PLANNING (in Caire)

TIME WINDOWS - Broad allowed periods

  • Example: "Morning 07:00-10:00" for daily visits, "Flexible 07:00-22:00" for weekly cleaning
  • Purpose: Find optimal FIXED TIMES for recurring patterns across all frequencies (daily, weekly, monthly, annually)
  • Result: "Every Monday 08:00-08:40" (daily) or "Every other Wednesday 14:00-15:30" (weekly) or "First Monday of each month 10:00-11:30" (monthly)

⬇️ Customer meeting & approval

🟢 PHASE 2: DAILY OPTIMIZATION (in Welfare/Epsilon/Carefox)

FLEXIBILITY MINUTES - Small buffers around fixed time

  • Example: "08:00 ±15 minutes" → can be performed 07:45-08:15
  • Purpose: Handle daily variations (traffic, delays) for all visit types
  • Result: Adjusted daily schedule within approved windows - applies to both daily and weekly/monthly visits when they are to be performed

Core Concepts

1. Time Window (Pre-Planning)

Definition: Allowed period during which a visit CAN be scheduled.

System handles all frequencies: Daily, weekly, bi-weekly, monthly, annually - based on municipal decision specifications.

Examples of time windows:

Examples of frequencies from municipal decisions:

Purpose: Guide AI optimization to find feasible patterns that match:

Should not be confused with: Flexibility minutes (set later for daily optimization)

2. Flexibility Minutes (Daily Optimization)

Definition: Small buffer (±15 or ±30 minutes) around a FIXED time. Applies to all visit types when they are to be performed - regardless of whether the visit is daily, weekly, monthly or annual.

Examples:

Purpose: Handle daily variations without replanning the entire week/month:

Set in: Welfare/Epsilon or Carefox (NOT in pre-planning phase)

3. Mandatorya vs Optionala Visit

Mandatory (Hard Requirements):

  • Must be scheduled
  • Example: medication, toilet visits, meals, bedtime
  • Cannot be skipped even during resource shortage

Optional (Soft Requirements):

  • Can be postponed or skipped during resource conflicts
  • Example: walks, certain cleaning tasks, social activities
  • Helps schedulers/coordinators (or equivalent roles) prioritize without compromising critical care

Usage: During optimization, mandatory visits are scheduled first; optional visits fill remaining capacity.

4. All Visits Can Be Pre-Planned

Insight: Since municipal decisions do not contain fixed times, ALL visits can be treated as flexible during pre-planning. Biggest impact: Weekly/bi-weekly flexible visits like cleaning, groceries, and social activities where AI can find optimal times across multiple weeks. Daily visits (breakfast, medication) are often already relatively fixed but can also be optimized for better route patterns.

Process:

  1. Receive municipal decisions: Upload PDF/documents in Caire (1 or 300 decisions - system handles batch upload)
  2. OCR extraction: Caire automatically extracts customer info, service types, duration, frequency and skill requirements via AI-driven OCR. Completely replaces manual entry.
  3. Review and adjust: Scheduler/coordinator (or equivalent role) reviews extracted data and adjusts if needed
  4. Assign time windows: System automatically suggests time windows based on service type (e.g. "Cleaning" → "Flexible 07:00-22:00", "Breakfast" → "Morning 07:00-10:00")
  5. Run pre-planning: AI optimizes all visits together - especially effective for weekly/bi-weekly visits where optimal times can vary
  6. Present to customer and finalize: Proposed schedules presented, customer approves
  7. Enter fixed times: Export to Welfare/Epsilon/Carefox with fixed times

No more "trial loop" – all visits optimized together from the start. PDF upload eliminates manual data entry.

Complete User Workflow

🎨 Se interactive UX mockups: Click here to see interface designs

Overview Diagram

flowchart TD A[Municipal Decision
PDF/Document] --> B[📝 MANUAL FLOW
Entry in Caire] A -.->|🤖 AUTO-MAGIC| AUTO[PDF Upload OCR
AI data extraction] B --> C[Assign time windows
per service type] C --> D[Set visit parameters
Duration, frequency, priority] D --> E[AI pre-planning
Comprehensive optimization] AUTO -.->|Skips manual steps| E E --> F[Review proposed schedule
Check feasibility] F --> G{Schedule
feasible?} G -->|No| H[Adjust parameters
Time windows, priorities] H --> E G -->|Yes| I[Customer Meeting
Present proposed times] I --> J{Customer
satisfied?} J -->|No| K[Enter preferences
Specific time requests] K --> L[Re-optimize with
preferences as requirements] L --> I J -->|Yes| M[Finalize in Caire
Document agreement date] M --> N[Manual entry in
Welfare/Epsilon or Carefox] N --> O[Set flexibility minutes
±15 or ±30 min per visit] O --> P[Send implementation plan
to municipality] P --> Q{Which
system?} Q -->|Carefox| R[Daily schedules imported
to Caire for optimization] Q -->|Welfare/Epsilon| S[Daily planning in
Welfare/Epsilon manually] R --> T[Follow-up pre-planning
Every 30-45 days] S --> T style AUTO fill:#2563EB,stroke:#059669,stroke-width:3px,color:#fff style B fill:#FEF3C7,stroke:#F59E0B,stroke-width:2px

Detailed Step-by-Step Process

💡 UX Mockups for each phase:

Phase 1: PDF OCR - Automated Data Entry

🤖 AI-Driven PDF Upload & OCR - Batch Processing

🎨 View UX mockup: PDF Upload

Powerful batch upload: Upload 1 or 300 municipal decisions simultaneously - system handles all automatically.

Auto-Magic Process:

  1. Batch Upload PDF: Upload one or multiple municipal decisions directly in Caire (supports batch upload of hundreds of documents)
  2. AI OCR Extraction: Intelligent extraction of all decisions in parallel:
    • Customer info (name, address, ID number)
    • Service types (breakfast, medication, shower, cleaning, groceries, etc.)
    • Duration (length in minutes)
    • Frequencies (daily, weekly, bi-weekly)
    • Skill requirements (delegation, language, etc.)
  3. Auto-Assign Time Windows: AI automatically assigns time windows based on service type:
    • Breakfast → Morning (07:00-10:00)
    • Lunch → Lunch (11:00-13:00)
    • Medication → Morning/Evening (depending on type)
    • Shower → Morning or Afternoon
    • Cleaning → Flexible (07:00-22:00) - weekly/bi-weekly visits with biggest optimization potential
    • Groceries → Flexible (07:00-22:00) - weekly visits
    • Social activities → Flexible - weekly visits
  4. Review & Approve: Scheduler/coordinator (or equivalent role) reviews extracted data and approves (or adjusts if needed)

Time Savings: From 30-60 minutes manual entry per customer → 2-5 minutes review per customer. With batch upload of 100 decisions: from 50-100 hours → 3-8 hours total.

Replaces manual entry: PDF upload is now the primary method. Manual entry still available for special cases or if PDF quality is poor.

Phase 2: Pre-Planning Optimization

Step 5: Run AI Pre-planning

🎨 View UX mockup: Steps 3-4

Scheduler/coordinator (or equivalent role) clicks "Run pre-planning" in Caire.

System actions:

  1. Fetches all visits within planning period (e.g. next 30 days)
  2. Fetches all staff schedules and skills
  3. Builds comprehensive optimization input
  4. Runs AI optimization
  5. Receives optimized schedule with proposed fixed times

Duration: 5-15 minutes for typical organization

Example result:

Anna Svensson - Breakfast (40 min)
✓ Mon: 08:00-08:40 (Staff: Lisa K.)
✓ Tue: 08:30-09:10 (Staff: Maria S.)
✓ Wed: 08:10-08:50 (Staff: Lisa K.)
✓ Thu: 07:55-08:35 (Staff: Johan A.)
✓ Fri: 08:20-09:00 (Staff: Maria S.)
✓ Sat: 08:00-08:40 (Staff: Lisa K.)
✓ Sun: 08:00-08:40 (Staff: Weekend staff)
Travel efficiency: Up to 20% reduction in total travel time
Staff utilization: 75-80% (target range)
All mandatory visits: ✓ Scheduled
Optional visits: 95% scheduled

Step 6: Scheduler/coordinator (or equivalent role) Reviews Schedule

Review for:

  • All mandatory visits scheduled
  • Staff skills match visit requirements
  • Trofel times reasonable
  • No schedule conflicts
  • Customer lifestyle considerations

If problems found → Adjust parameters and re-run (Step 5)

Phase 3: Customer Engagement

Step 7: Customer Meeting (Mandatoryt)

🎨 View UX mockup: Step 5-6

Purpose: Present proposed schedule and agree on fixed times.

Process:

  1. Book meeting with customer (in-person or phone)
  2. Present the proposed week pattern from Caire
  3. Explain reasoning (e.g. "We scheduled your breakfast at 08:00 because your caregiver Lisa has other customers in the area in the morning")
  4. Ask if the times fit customer's routine

Two Outcomes:

A) Customer Satisfied (Typical Case - 70-80% of cases)

  • Customer confirms proposed times work
  • Go to Step 10 (Finalize)

B) Customer Has Specific Preferences (20-30% of cases)

  • Customer requests specific changes
  • Go to Step 8 (Preferences)

Step 8: Enter Customer Preferences (Optional)

Only if customer is not satisfied with initial proposal.

Example preferences:

  • "Breakfast must be 08:30, not 08:00"
  • "Shower must be Monday/Wednesday, not Tuesday/Thursday"
  • "Prefers female caregiver for showering"
  • "No visits before 09:00 on weekends"

How to enter in Caire:

  • Mark visit as "Has customer preference"
  • Enter specific time or narrower window
  • Add preference notes

Step 9: Re-optimize with Preferences

System treats preferences as hard requirements:

  • Fixed times become locked
  • Narrow time windows applied
  • Skills/gender preferences prioritized

Re-run optimization → New proposal considering preferences

Important: May require compromises:

  • Increased trofel time
  • Different staff assignments
  • Some optional visits may need to be postponed

Present new proposal to customer (return to Step 7)

Step 10: Finalize in Caire

🎨 View UX mockup: Steps 7-8

When customer approves:

  1. Click "Finalize schema" i Caire
  2. System registers:
    • Finalizeda fixed times for varje visits
    • Agreement date
    • Assigned staff
    • Notes from customer meeting
  3. Status changes to "Approved"
  4. Schedule locked for export

Phase 4: Integration with Operations Systems

Step 11: Manual Entry in Welfare/Epsilon or Carefox

Scheduler/coordinator (or equivalent role) manually enters each visit in operations system:

For Welfare/Epsilon:

  • Enter customer info (if new)
  • Create visit entry
  • Set fixed time (e.g. "Monday 08:00-08:40")
  • Assign staff
  • Critical: Set flexibility minutes (e.g. ±15 min)
  • Repeat for each visit in week pattern

For Carefox:

  • Enter customer info (if new)
  • Create visit entry
  • Set fixed tid
  • Assign staff
  • Set flexibility minutes
  • Upprepa for varje visits

Step 12: Set Flexibility Minutes

Purpose: Allow daily optimization within small buffer.

Typical Values:

  • ±15 minutes: Standard for de flesta visits
  • ±30 minutes: Flexible visits (e.g. cleaning)
  • ±0 minutes: Time-critical (e.g. medication must be exact)

Example:

Fixed tid: 08:00-08:40
Flexibilitet: ±15 min
Actual execution window: 07:45-08:55

Denna flexibilitet uses of daily optimizer to handle:

  • Traffic
  • Delays from previous visits
  • Customer not ready
  • Staff started shift late

Step 13: Send Implementation Plan to Municipality

Implementation plan is sent to the municipality for review and approval according to the municipality's process and system.

Process:

  1. Generate implementation plan document from Welfare/Epsilon/Carefox
  2. Upload to municipality's system (each municipality has its own system for implementation plan documentation)
  3. Send to municipality for review
  4. Customer signs implementation plan

Important: Different municipalities have different systems and requirements for implementation plan documentation.

Phase 5: Automated Nightly Operations

Step 14: Nightly Auto-Synchronization (02:00 AM)

🌙 Automated process while you sleep:

graph LR A[🌙 02:00 AM
Batch Job Starts] --> B[📥 Auto-Import
60-day rolling window] B --> C[🔄 Smart Sync
Detect changes] C --> D[🤖 AI Optimization
Pre-defined scenarios] D --> E[☀️ Morning
Results Ready]

Import & Synchronization:

  • Auto-import: Fetches new/updated schedules from Carefox (60-day rolling window)
  • Change Detection: Identifies new visits, updated times, cancellations
  • Intelligent Merge: Applies changes without overwriting existing data
  • Conflict Handling: Automatic resolution of minor conflicts, flags major ones for manual review

Automatic AI Optimization:

  • Pre-defined scenarios: Runs all configured scenarios automatically
  • Daily optimization: Optimizes today's schedules with flexibility minutes (±15/±30 min)
  • Monthly planning: Runs pre-planning for movable visits across entire month (Ghost Ideal Schedule)
  • Results ready in morning: Fresh KPIs, optimization insights, and comparisons

Time Savings: Zero manual work - everything happens automatically while you sleep

Integration per System
System Import Optimization Export
Carefox ✅ Auto-import nightly (API) ✅ AI optimization automatic ⚠️ Manual export (API in development)
Alfa eCare Welfare 🔜 Roadmap (API integration planned) ✅ Can run manually ⚠️ Manual export

Phase 6: Continuous Improvement

Step 15: Monthly Pre-Planning Re-run

Frequency: Automatic monthly (included in nightly batch)

Ghost Ideal Schedule Concept:

  • Ideal Baseline: Monthly optimization of all recurring visits with flexible time windows
  • Daily Reality: Operational schedules handling real-world constraints (sick calls, urgent visits)
  • Efficiency Tracking: Compare daily operational efficiency vs ideal baseline
  • Recommendations: System suggests when it's time to re-plan movable visits

Triggers for manual re-planning:

  • Efficiency drops below 75% (from 80%+ ideal)
  • New customers added affecting route patterns
  • Significant staff changes
  • Customer requests for schedule changes

Benefits:

  • Continuous optimization automatically in background
  • Data-driven decisions on when re-planning is needed
  • Measures real efficiency loss from daily disruptions

System Architecture

Component Diagram

graph TB subgraph Kommun[Kommun] M[Municipal Decision
PDF/Dokument] end subgraph Caire[Caire-plattform] PE[Pre-planning engine] AI[AI-Optimization] VS[Visitslager
Time Window & parametrar] FS[Finalizeda Scheman
Approveda pattern] end subgraph Operations[Verksamhetssystem] WE[Welfare/Epsilon] CF[Carefox] end M -->|Manual entry| VS VS -->|All visit + personal| PE PE -->|Optimization request| AI AI -->|Beforeslagt schema| PE PE -->|Review & approve| FS FS -->|Manual entry
Fixed times + flex| WE FS -->|Manual entry
Fixed times + flex| CF CF -->|Daily import| PE PE -->|Daily optimization| CF style Caire fill:#e1f5ff style Operations fill:#fff4e1

Time Windows vs Flexibility

Visual Comparison

gantt title Time Window (Pre-Planning) vs Flexibility Minutes (Daily) dateFormat HH:mm axisFormat %H:%M section Pre-Planning Time Window (Allowed Period) :crit, 07:00, 3h AI-Beforeslagen Fixed Tid :milestone, 08:00, 0h section Daily Drift Fixed Tid (Bas) :active, 08:00, 40m Flexibilitet ±15min :done, 07:45, 1h10m Actual Execution :milestone, 08:10, 0h

Decision Matrix: When To Use Each

Scenario Phase Verktyg Flexibilitetstyp Examples
Initial planning for ny customer Pre-Planning Time Window Broad (e.g. 3-hour window) "Frukost can vara 07:00-10:00"
Customer Meeting Pre-Planning Fixed Tidsproposal Ingen (specifika tider) "Vi suggests 08:00-08:40"
Enter i verksamhetssystem Handover Fixed Tid + Flex Liten buffert (±15 min) "08:00 with ±15 min flexibilitet"
Daily execution Daily Drift Adjustd Tid Within flex-buffert "Start 08:10 due to traffic"
Traffic delay Daily Drift Realtidsjustering Within flex-buffert "Delayed to 08:25"
Customer requests permanent change Omplanning Nytt time windows Broad or narrow based on request "Customer vill alltid ha 09:00"

Examples

Example 1: Dailyt Visit with Enkelt Time Window

Municipal Decision:

  • Insats: Frukosthelp
  • Duration: 40 minuter
  • Frequency: Dailyen (7 dagar/vecka)
  • Rekommenderad tid: "Runt 08:30"

Entry in Caire:

  • Title: Frukost
  • Duration: 40 min
  • Frequency: Veckovis
  • Dagar: 7 (alla dagar)
  • Time Window: Morning (07:00-10:00)
  • Prioritet: Mandatory
  • Required skills: None

AI-proposal (Pre-Planning):

Mon: 08:00-08:40 (Personal: Lisa K., Travel: 12 min from previous)
Tis: 08:30-09:10 (Personal: Maria S., Travel: 8 min from previous)
Ons: 08:10-08:50 (Personal: Lisa K., Travel: 10 min from previous)
Tor: 07:55-08:35 (Personal: Johan A., Travel: 15 min from office)
Fre: 08:20-09:00 (Personal: Maria S., Travel: 7 min from previous)
Sat: 08:00-08:40 (Personal: Lisa K., Travel: 12 min from previous)
Sun: 08:00-08:40 (Personal: Weekend staff, Travel: 20 min from office)

Customer Meeting:

  • Scheduler/coordinator (or equivalent role): "Vi suggests 08:00-08:40 de flesta dagar. Passar det?"
  • Customer: "Yes, perfect. I usually wake up at 07:30."

Finalizet i Caire:

  • Approved: Yes
  • Agreement date: 2025-10-16

Entry in Welfare/Epsilon:

  • Fixed times enligt ovan
  • Flexibilitet: ±15 minuter per visits
  • Result: Can be performed 07:45-08:55 daily as needed

Example 2: Varannan Veckas Visit with Flerveckors Time Window

Municipal Decision:

  • Insats: Cleaning
  • Duration: 90 minuter
  • Frequency: Varannan vecka
  • Rekommenderad tid: "Flexibel, optional tid"

Entry in Caire:

  • Title: Cleaning
  • Duration: 90 min
  • Frequency: Varannan vecka
  • Time Window: 2-week window (13 okt 07:00 - 26 okt 22:00)
  • Prioritet: Optional
  • Required skills: None

AI-proposal:

Vald slot: Wednesday 15 okt, 14:00-15:30
Motivering:
- Personal Maria S. have lucka between visit
- Close to other clients in same building
- Fits customer schedule (home from work)

Customer Meeting:

  • Scheduler/coordinator (or equivalent role): "Vi suggests Wednesdaysaftermiddagar varannan vecka kl 14:00"
  • Customer: "Perfekt"

Finalizet:

  • Fixed pattern: Varannan Wednesday 14:00-15:30
  • Flexibilitet: ±30 minuter (flexibel insats)

Example 3: Mandatoryt Medicineringsvisit

Municipal Decision:

  • Insats: Medicintillsyn
  • Duration: 15 minuter
  • Frequency: Dailyen, 2 times per day (morning and evening)
  • Recommended times: "08:00 and 20:00"
  • Requires delegation

AI-proposal:

Morgon: 08:00-08:15 (Personal: Lisa K. - has delegation)
Evening: 20:00-20:15 (Personal: Johan A. - has delegation)
Motivering:
- Tider matches rekommenderade tider from kommun
- Both personal have necessary delegation
- fit within otherg visits scheman

Customer Meeting:

  • Scheduler/coordinator (or equivalent role): "Medicinering must vara vid specific tider. Vi suggests 08:00 and 20:00"
  • Customer: "Yes, my medication schedule requires these times"

Finalizet:

  • Morning: 08:00-08:15, Flexibilitet: ±0 minuter (tidskritiskt)
  • Evening: 20:00-20:15, Flexibilitet: ±0 minuter (tidskritiskt)

Important: Ingen flexibilitet for medicineringsvisit - must vara exakta tider.

How Pre-Planning Transforms Home Care Scheduling

Pre-Planning with Caire delivers transformation through:

  1. Treat ALL visits as flexible under initial planning (since municipal decisions do not have fixed times)
  2. Use broad time windows to guide AI optimization
  3. Find optimal fixed times that balances travel, staff skills and customer lifestyle
  4. Engage customers with feasible proposals (inte orealistiska manuella gissningar)
  5. Finalize schedules with customer approval before entry i verksamhetssystem
  6. Set flexibility minutes i Welfare/Epsilon/Carefox for dailya adjustments
  7. Enable daily optimization (for Carefox customers) within approved buffers
  8. Support periodic re-planning to continuously improve as organization evolves

Skillnad:

  • Pre-Planning = Find the pattern (breda time windows → fixed times)
  • Daily Optimization = Execute the pattern (small flexibility around fixed times)

Result: Over 50% time savings during introduction, improved billing efficiency (75-80%+ service time target), higher customer satisfaction, better staff utilization.