Fully Reloaded introduced an AI-powered chatbot to its WhatsApp communication strategy to streamline customer support and reduce the load on its support team. In this case study, we’ll walk through how an AI WhatsApp chatbot helped Fully Reloaded automate the majority of incoming inquiries and deliver faster, more consistent customer service without expanding the team.
Project overview
Channel: WhatsApp
Platform: SendPulse
Implementation partner: WiseFlow, a chatbot development and LMS integration platform
Current status: live 24/7, automatically handling customer support inquiries
About the client
Fully Reloaded is a Germany-based rental service that offers portable charging solutions across Berlin. The company rents out power banks for smartphones, headphones, smartwatches, and other mobile devices, helping people stay connected while moving around the city.
Why Fully Reloaded needed a chatbot
Fully Reloaded’s target customers are young, active individuals and business professionals who are always on the go. For this audience, speed, convenience, and seamless service are essential.
However, before implementing the chatbot, the company faced several challenges in meeting these expectations consistently.
Long response times
Fully Reloaded’s support team was handling a high volume of incoming requests every day. This created a heavy workload for support agents and led to longer response times for customers.
At the same time, customers contacted support around the clock, often with urgent issues such as a malfunctioning power bank or a device return. Since agents worked fixed hours, immediate responses weren’t always possible.
For an audience that expects solutions here and now, delayed responses negatively affected the customer experience and increased the risk of losing user trust and loyalty.
Unnecessary workload on the customer support team
A significant portion of support inquiries involved repetitive questions or common technical issues. As a result, agents spent valuable time answering the same queries instead of focusing on more complex cases requiring human expertise.
This imbalance in workload not only reduced team efficiency but also made it difficult to maintain a consistently high level of service.
Lack of structured customer data
Customers often described their issues in free-form messages and frequently left out essential details such as device model or rental location.
This led to fragmented, unstructured data that made it challenging for support agents to assess the situation quickly. As a result, agents had to ask multiple follow-up questions, which further prolonged the resolution time and slowed down the entire support process.
How the chatbot improved Fully Reloaded’s support operations
The chatbot built with SendPulse operates directly on WhatsApp, the same channel customers were already using to reach support. This eliminated any need for users to adjust to a new platform or communication flow, making the transition seamless.
Powered by automation and AI, the chatbot efficiently handles routine workflows, boosting efficiency and customer satisfaction. If you want to replicate this approach, WhatsApp customer support software helps automate FAQs, keep conversations organized, and smoothly hand chats over to agents when needed.
Better customer experience
Unlike human agents, the chatbot responds within just five seconds, works 24/7, and can handle an unlimited number of conversations simultaneously. It also automatically detects the language of each message and provides support in multiple languages, allowing Fully Reloaded to serve a wider audience without any additional effort.
At the same time, the bot communicates in a natural and friendly tone, making interactions feel more like a conversation with a real person than a rigid automated system. This combination of speed, availability, and human-like communication significantly improves the overall customer experience and strengthens brand loyalty.
Reduced workload for support agents
The chatbot now manages routine inquiries related to power bank rentals, charging station locations, app downloads, and frequently asked questions about services. This shift ensures that customers get instant responses, while support agents are no longer overloaded with repetitive tasks.
With clear AI logic and instructions, the bot can analyze support requests and resolve common technical issues faced by renters. If a request is too complex or lacks sufficient information, the bot seamlessly transfers the conversation to a human agent.
This approach ensures that only cases requiring personal attention reach support agents, allowing them to focus on high-priority issues and improving overall team efficiency.
Structured data collection for faster resolution
If the chatbot can’t resolve an issue on its own, it collects full context and generates a structured support request for the agent. Notifications about new requests are instantly sent to the internal Telegram group via SendPulse Notifications and include key information such as:
- rental location,
- station code,
- device model,
- problem description.
This allows support agents to immediately understand the nature and complexity of the issue and join the conversation fully prepared.
The chatbot flow in detail
The WiseFlow team built a “Standard reply” flow for Fully Reloaded’s WhatsApp chatbot.
When users access the bot via a link in the mobile app or scan a QR code at power bank rental or return stations, the first step is a “Filter” element. This block checks if a greeting message has already been sent for the day. If not, the user is promptly greeted with a static welcome message from Fully Reloaded’s support team.
The “Standard reply” сhatbot flow
Once the user asks a question or describes an issue, the conversation moves to the “AI Agent” element. There, artificial intelligence analyzes the user’s request and continues the conversation based on predefined instructions and the element’s technical settings.
Next, we’ll take a closer look at how these settings shape the chatbot’s behavior and how it generates responses.
AI prompt logic
The AI instruction includes all essential information about power bank rental terms, current pricing, return rules, and available payment methods. This allows the chatbot to provide accurate and relevant answers to both new users exploring the service and existing customers who urgently need a power bank.
Common support scenarios are defined in advance, such as issues with malfunctioning power banks, stations that fail to dispense or accept devices, or deposits that haven’t been refunded. For each scenario, the AI follows a clear action plan: it knows what guidance to offer, which follow-up questions to ask, and what data to collect.
AI instruction logic
This structured approach allows the chatbot to independently handle most support requests without involving a human agent. However, the AI is programmed with limitations: if it lacks enough information to provide an accurate answer, it will transfer the conversation to a support agent.
The AI is also restricted from answering unrelated or personal questions, sharing the company’s phone number, apologizing on behalf of the company, or promising compensation. These safeguards ensure that the chatbot communicates accurately, maintains professionalism, and upholds Fully Reloaded’s brand integrity.
Technical setup
The AI Agent was configured with the following technical parameters:
| AI model |
GPT-4.1 mini |
| Context window |
the last 15 messages |
| Maximum response length |
500 tokens (approximately 375 words) |
| Temperature (creativity level) |
0.2 |
| Execution mode |
conditional exit |
To ensure seamless customer service, the “AI Agent” block has predefined exit conditions. When any of these conditions are met, the chatbot stops generating AI responses and immediately hands the conversation over to a human support agent. This ensures that uncommon or complex situations are always handled individually by a qualified specialist.
The chatbot exits the AI flow in the following cases:
- the user sends a photo, video, image, or link;
- a power bank rental or return station is malfunctioning;
- the user explicitly requests help from a human agent;
- the user asks for an invoice;
- the user can’t return the device due to being in a different city;
- a callback request is made in exceptional cases.
Additionally, the chatbot exits the AI flow once it has gathered all the necessary information to create a structured support request.
In all these scenarios, the chatbot immediately notifies the user that their request has been received and will be processed shortly. At the same time, a notification is sent to the internal Telegram group, alerting a support agent. This notification includes all the data collected by the bot, ensuring a quick and efficient resolution by the support team.
Results achieved with the AI-powered chatbot
By integrating the WhatsApp chatbot, Fully Reloaded achieved significant improvements across its customer support operations. Here’s how the chatbot transformed their workflows:
- Automated 60-70% of incoming support requests, allowing agents to focus on more complex and unique cases.
- Simplified issue analysis and faster resolution by collecting structured support requests through the chatbot.
- Reduced average response time during business hours from 10 minutes to just 5 seconds.
- Enabled 24/7 customer support without expanding the team or increasing workload for agents.
- Strengthened brand perception, with customers now describing the service as convenient, modern, and easy to use.
Fully Reloaded’s experience demonstrates that automating routine workflows can provide a competitive edge. Customers benefit from faster, more reliable support, while teams have the time and capacity to address more complex issues and deliver higher-value service.
Ready to boost your support efficiency? Build your own AI-powered chatbot with SendPulse, or get assistance from certified partners and free your team to focus on tasks that truly require human expertise.