How to Configure Hours of Operation with After Hours AI Response
Your sales team works from 9 to 6. Your customers are messaging you at 10 p.m. This mismatch between your business hours and the actual behavior of your prospects is costing you sales every day. The solution isn't hiring night shifts or asking your team to work from home: it's setting up a hybrid system where human agents and artificial intelligence work together, each contributing the most value during their peak hours.
In this tutorial, we explain step by step how to configure business hours on your WhatsApp Business platform, how to define AI autonomy levels according to the schedule, and how to ensure your business offers 24/7 support without sacrificing the quality of human interaction when available.
Why You Need a Human + AI Hybrid System
The traditional model is broken.
Most businesses operate under one of two extremes: either they rely entirely on human agents (limited to office hours), or they activate a generic chatbot that gives the same robotic response regardless of whether it's Monday at 10am or Sunday at midnight.
Both extremes have serious problems:
- Humans only: You lose the 60% number for inquiries received outside of business hours. Your agents are overwhelmed in the morning responding to messages accumulated from the previous day.
- Basic chatbot only: The experience is frustrating during working hours, when a human agent could resolve the query with greater empathy and negotiation skills.
The hybrid model: the best of both worlds
A smart hybrid system works like this:
During business hours:
- The AI receives the first message and qualifies the prospect
- It identifies the customer's intent and routes them to the most appropriate agent.
- While the agent is on the call, the AI suggests responses and provides context.
- If all agents are busy, the AI keeps the conversation active until one becomes available.
Outside of business hours:
- The AI provides completely autonomous service
- Answer frequently asked questions with accurate business information
- Qualify leads by collecting key prospect data
- Schedule appointments for the next business day
- Escalate real emergencies to a designated contact
This model guarantees that no message goes unanswered, no matter what time it arrives.
Step 1: Define your Daily Business Hours
The first step is to configure the times when your human team is available to handle conversations. This configuration forms the basis on which the system will decide when the AI will act autonomously and when it will function as an assistant to your agents.
Setting schedules in Aurora Inbox
In Aurora Inbox, schedule configuration is done from the administration panel:
- Navigate to Settings > Business Hours
- Define the schedule for each day of the week independently.
- You can set up multiple blocks per day (for example, morning and afternoon with a lunch break)
Example of a typical configuration:
| Day | Hours of operation | Pause |
|---|---|---|
| Monday to Friday | 9:00 – 18:00 | 14:00 – 15:00 |
| Saturday | 9:00 – 14:00 | – |
| Sunday | Closed | – |
Considerations for setting schedules
Analyze when your customers write: Before setting up schedules, review your incoming message statistics. You might find that 40% errors from your inquiries are arriving between 6 PM and 9 PM. In that case, it might be worthwhile to extend your human support hours until 8 PM instead of relying solely on AI.
Includes time zone: If you serve clients in multiple countries, configure your team's reference time zone. Aurora Inbox allows you to define a primary time zone, and the system automatically adjusts the AI's schedules accordingly.
Special hours and holidays: Configure exceptions for holidays, vacations, or special events when your team will be unavailable. During these periods, the AI operates in fully autonomous mode.
Step 2: Configure AI Autonomy Levels by Schedule
The key to an effective hybrid system is that the AI doesn't behave the same way at all times. During business hours, it should act as an assistant; outside of business hours, it should act as an autonomous agent.
Level 1: Assistant Mode (During business hours)
In this mode, AI fulfills a support role for your human agents:
Functions of AI as an assistant:
- First intelligent contact: When a customer writes, the AI sends a personalized greeting and asks initial qualification questions (name, which service you are interested in, urgency) while notifying the available agent.
- Intelligent routing: Based on the detected intent, the AI directs the conversation to the most appropriate agent. If you have agents specializing in sales, support, and collections, the AI identifies the topic and assigns it correctly.
- Real-time suggestions: While the agent is on duty, AI suggests responses based on customer history, relevant product information, and best service practices.
- Queue management: If all agents are busy, the AI informs the customer of the estimated waiting time and keeps the conversation active with useful information until an agent becomes available.
Aurora Inbox settings:
Mode: Assistant Schedule: Within configured business hours Allowed actions: - Initial greeting and rating - Routing to agents - Response suggestions - Reporting wait time - Answering basic FAQs while the customer waits Restricted actions: - Does not close sales without supervision - Does not schedule appointments without agent confirmation - Does not provide discounts or special conditions
Level 2: Autonomous Mode (Outside of business hours)
Outside of regular hours, AI becomes a full-fledged agent capable of resolving conversations from start to finish:
AI functions in autonomous mode:
- Full consultation support: It answers questions about products, services, prices, locations, and any information you have trained in its knowledge base.
- Lead qualification: Collect prospect data (name, company, budget, need, urgency) and classify them according to your qualification criteria.
- Appointment scheduling: Check availability on your calendar and offer times for the next business day. The client confirms and the appointment is automatically registered.
- Sending information: Share catalogs, price lists, technical data sheets, locations, and any relevant sales materials based on the query.
- Emergency escalation: Identify situations that require immediate human attention and contact the designated person in charge.
Aurora Inbox settings:
Mode: Autonomous Schedule: Outside of configured business hours Permitted actions: - Answer all knowledge base queries - Qualify leads with completed forms - Schedule appointments during available times on the next business day - Send catalogs and sales materials - Register quote requests - Collect information for follow-up Restricted actions: - Does not offer unauthorized discounts - Does not confirm orders requiring manual review - Does not share other customers' confidential information Automatic escalation: - Medical emergencies (clinics/hospitals) - Serious complaints with threat of cancellation - Requests that mention urgent keywords
Step 3: Configure Transition Messages (Handoff)
The transition moments between AI and human (and vice versa) are critical to the customer experience. A poorly executed handoff generates confusion and frustration.
AI to Human Transition (start of schedule)
When business hours begin and there are ongoing conversations that the AI was handling overnight:
Suggested transition message:
"Good morning, [customer name]. I'm [agent name] from the [company] team. I saw that you were talking to our virtual assistant last night about [topic]. I now have all the context of your inquiry. How can I help you now?"
Key configuration:
- The agent receives an automated summary of the nighttime conversation.
- AI transfers the full context: customer data, intent, questions asked, shared information
- The customer does not need to repeat information
Human to AI Transition (end of schedule)
When business hours end and there are active conversations:
Option A – Conversation almost resolved:
"It was a pleasure assisting you, [name]. From now on, our virtual assistant can help you with any further questions. If you need anything else, please don't hesitate to write. A team member will follow up first thing tomorrow if necessary."
Option B – Conversation in the sales process:
"[Name], our in-person service hours have ended for today, but our virtual assistant can continue to help you with additional information, answer your questions, and if you wish, schedule an appointment for tomorrow so we can finalize everything with personalized attention."
Aurora Inbox settings:
- Navigate to Settings > Transition Messages
- Define messages for each type of transition
- Customize the available variables: customer name, agent name, conversation topic
- Activate automatic summarization so the agent receives context the next day
Step 4: Configure Emergency Escalation Rules
Not all inquiries can wait until the next business day. Some situations require immediate human attention, regardless of the time.
Define emergency criteria
Configure keywords, phrases, or conditions that trigger immediate escalation:
Examples by industry:
- Medical clinics: "severe pain", "urgency", "emergency", "bleeding", "accident"
- Real estate: "I want to move today", "problem with the apartment", "water leak"
- E-commerce: "Incorrect order already shipped", "Duplicate charge", "Card fraud"
- Professional services: "Hearing tomorrow", "Deadline expires today", "Urgent lawsuit"
Configure the escalation flow
Escalation Rules: High Priority: - Keywords: [emergency keyword list] - Action: Notify the on-duty manager via SMS and call - Maximum wait time: 5 minutes - Message to the customer: "Your case requires immediate attention. I am contacting a specialist who will assist you in the next few minutes." Medium Priority: - Condition: Existing customer with a serious complaint - Action: Notify the supervisor via push notification - Maximum wait time: 30 minutes - Message to the customer: "I understand the urgency of your situation. I have escalated your case so that a specialist will contact you as soon as possible." Low Priority: - Condition: Inquiry beyond AI capacity - Action: Register for priority attention at the start of business hours - Message to the customer: "Your inquiry requires review by a specialist. I have marked it as a priority so that you will be seen first thing tomorrow."
Configure on-call contacts
In Aurora Inbox you can designate emergency contacts with rotation:
- Navigate to Settings > Escalation > On-Call Contacts
- Add team members who can receive after-hours alerts
- Define a rotation schedule (for example, each agent covers one week)
- Configure your preferred notification method: SMS, call, email, or push notification
Step 5: Train the AI for Autonomous Operation
For AI to function effectively outside of business hours, it needs access to all the information that a human agent would use to serve customers.
Essential information for the knowledge base
- Full catalog of products or services with prices, features and availability
- Frequently Asked Questions organized by topic and with detailed answers
- Policies warranty, returns, cancellations and exchanges
- Locations with directions, schedules and arrival information
- Purchase process step by step for each product or service
- Requirements documentation, preparation, or prerequisites for appointments
- Current promotions with conditions, validity and restrictions
Configure the lead qualification form
Define what information AI should collect when identifying a qualified prospect:
Qualification fields: Required: - Full name - Service/product of interest - Preferred contact method Optional (depending on industry): - Estimated budget - Desired service date - Number of people/units - Client location Qualification criteria: - Hot lead: Has a budget, immediate need, decision in less than 1 week - Warm lead: Interested but comparing options, decision in 2-4 weeks - Cold lead: Researching, no urgency, decision in more than 1 month
Expected Results of the Hybrid Model
When you correctly implement the hybrid human + AI coverage scheduling configuration, the typical results are:
| Metrica | Formerly | Then |
|---|---|---|
| Unanswered messages outside of business hours | 100% | 0% |
| Average first response time | 45 min – 12 hours | Less than 30 seconds |
| Qualified leads at the start of the day | 0 | 8-15 leads ready for follow-up |
| Appointments scheduled outside of business hours | 0 | 3-7 appointments daily |
| Customer Satisfaction | 65% | 92% |
| Team's morning workload | High (accumulated messages) | Baja (pre-qualified leads) |
The most significant impact is that your human agents start their day with pre-qualified leads, scheduled appointments, and complete context for each conversation. Instead of wasting the first hour of the day responding to 30 accumulated messages with "hello, how can I help you?", they can focus directly on closing sales with prospects that the AI prepared overnight.
Best Practices for Configuration
1. Start conservatively and gradually increase autonomy
Configure the AI with limited permissions for the first two weeks. Review the nightly conversations each morning, identify areas for improvement, and gradually expand the allowed actions as you gain confidence in the responses.
2. Review autonomous conversations daily
For the first month, dedicate 15 minutes each morning to reviewing the conversations that the AI handled autonomously. This allows you to identify questions that need more information, detect errors, and continuously improve.
3. Update the knowledge base regularly
Every time your human team answers a new question that the AI couldn't handle, add that information to the knowledge base. In just a few weeks, the AI will be handling 951% of queries without assistance.
4. Customize the transition messages
Avoid generic messages like "an agent will assist you." Include the customer's name, reference the topic of their inquiry, and provide value in each transition message.
5. Configure performance alerts
It sets up notifications for when AI cannot resolve a query, when a customer shows frustration, or when an important sales opportunity is detected that deserves immediate human attention, even outside of business hours.
Frequent questions
1. Does the customer know when they are speaking with AI and when with a human?
Yes, transparency is key. Aurora Inbox allows you to configure clear identifiers. When AI is responding, it can introduce itself as "virtual assistant from [your company]." When a human agent takes over, they introduce themselves using their real name. Studies show that customers have no problem interacting with AI as long as it's honest, helpful, and efficient. What frustrates customers isn't talking to AI, but talking to an AI that doesn't resolve anything.
2. What happens if the AI doesn't know how to answer a question outside of business hours?
When the AI detects that it doesn't have enough information to provide an accurate answer, it has several configurable options: it can inform the customer that a specialist will follow up first thing in the morning, it can offer related information that it does have available, or it can escalate the inquiry if it meets urgency criteria. It never fabricates answers or provides incorrect information. In Aurora Inbox, you can configure exactly how the AI should behave when faced with questions that are outside its knowledge base.
3. Can I set different schedules for different departments or lines of business?
Yes, Aurora Inbox allows you to configure separate schedules for each support line or team. For example, you can have your sales team available from 8 a.m. to 8 p.m. with autonomous AI the rest of the day, while your technical support team operates from 9 a.m. to 6 p.m. with autonomous AI outside of those hours. Each team can have its own knowledge base, escalation rules, and AI autonomy levels.
4. How does the scheduling configuration affect the response times reported in my metrics?
Aurora Inbox separates metrics by time of day, allowing you to analyze the performance of your human and AI agents independently. You can see the average response time during business hours (team performance), the response time outside of business hours (AI performance), the autonomous resolution rate, the number of escalations, and customer satisfaction in each time slot. This allows you to optimize both support channels based on real data.
5. Can the AI's behavior be changed for special days like Black Friday or product launches?
Absolutely. Aurora Inbox allows you to create behavioral profiles that you can activate for special events. For example, during Black Friday, you can configure the AI in autonomous mode to have access to exclusive promotions, automatically apply discount codes, and have a lower threshold for escalating to a human agent on call. You can schedule these profiles in advance to activate and deactivate on specific dates and times without manual intervention.
Hybrid coverage scheduling is one of the most impactful features for businesses that receive inquiries outside of their regular business hours. With Aurora Inbox, you can implement this system in under an hour and start capturing leads, scheduling appointments, and resolving inquiries 24/7 without increasing your team or sacrificing the quality of service when your human agents are available.
