Skip to main content
To the homepage of Knowit

CASE

PLUXEE

The introduction of an AI-based chatbot significantly improved customer service for Pluxee

RETAIL & SERVICES

With surging demand for Pluxee's services came an increased workload on the support team. By introducing an AI-based chatbot, Pluxee was able to handle the preassure and, for the first time, consistently meet its SLA targets within the on-call support team.

    AI
    Mobile development

BACKGROUND

A global player in Employee Benefits and Engagement

Pluxee is a global player in Employee Benefits and Engagement that operates in 28 countries. Pluxee helps companies attract, engage, and retain talent thanks to a broad range of solutions across Meal & Food, Well-‏being, Lifestyle, Reward & Recognition, and Public Benefits. 

Knowit has had a long and successful collaboration with Pluxee and has been Pluxee's technology partner since 2019.


CHALLENGE

Boosted demand lead to increased pressure on the support team

The demand for Pluxee's services has increased sharply. But as usage increased, so did the demand for fast and accessible support across common service scenarios.

A significant share of incoming requests was related to a small number of recurring topics, placing pressure on support teams and increasing operational costs, particularly during peak seasons when traffic could increase up to four times.

More than 70% of incoming inquiries were related to a limited set of repetitive topics, such as card usage, account activation, and PIN management.

This created challenges in:

  • Scaling support during seasonal peaks
  • Maintaining fast response times
  • Ensuring consistent and accurate answers
  • Supporting a multilingual user base

Traditional chatbot solutions based on fixed decision trees were not flexible enough to handle the variety of ways users phrase their questions.


SOLUTION

An AI-powered conversational assistant

Knowit implemented an AI-powered conversational assistant integrated into Pluxee’s customer-facing application.

The solution combines large language model (LLM) capabilities with a retrieval augmented generation (RAG) approach, enabling the chatbot to interpret natural language while grounding responses in verified business content.

This approach ensures flexibility in understanding user intent while maintaining high precision through controlled knowledge sources. The system was built following a Security by Design approach, ensuring that security, compliance, and data protection were embedded from the start.

The implementation included:

  • Integration across web and mobile platforms
  • Backend development and system integration
  • Design of conversation flows and interaction logic
  • Multilingual support, including Polish, English, and Ukrainian
  • Built-in security mechanisms aligned with internal security and legal requirements 

In addition to answering questions, the chatbot supports self-service actions such as balance checks, PIN management, and acceptance point lookup, reducing the number of steps required to resolve user issues. The solution also provides an omnichannel experience, maintaining real-time conversation context across web and mobile interactions.

The system was designed for scalability, with an architecture that allows updates to the knowledge base within minutes and ensures consistent performance under varying load conditions.

The AI chatbot in a mobile device

RESULT

The solution has significantly improved the customer service

The solution is now a central part of Pluxee’s digital support experience.

Key outcomes include:

  • Handles around 250 conversations per day, each representing approximately four minutes of traditional support time
  • Resolves approximately 65% of interactions without escalation
  • Delivers efficiency equivalent to more than two full-time employees (FTEs) handling FAQ-related inquiries
  • Provides 24/7 support, essential during peak periods with up to four times higher traffic
  • Handled 821 questions across 391 users in a single day
  • Serves as a first point of contact, reducing pressure on customer support teams

Introducing the AI-based chatbot allowed us, for the first time, to consistently meet our SLA targets within the on-call support team.


Maciej Jędrak, New Digital Product Development Manager

Pluxee

"Introducing the AI-based chatbot allowed us, for the first time, to consistently meet our SLA targets within the on-call support team—especially during peak holiday periods in Poland, where this had always been a challenge.

Beyond delivering clear operational cost savings, the solution has significantly improved the overall level of customer service by providing faster and more accessible support. Additionally, the data we now collect gives us valuable insights into customer needs.

We’ve discovered that many inquiries go beyond the application itself—and thanks to the chatbot, we are now able to support those as well", says Maciej Jędrak, New Digital Product Development Manager at Pluxee. 

Do you want to know how we can assist you?

Give us a call or send a message, and we will get in touch with you. 

Tomasz Woropaj

Project Manager


Read more about our related offerings

Artificial Intelligence (AI)Mobile & app development