27% automatically processed customer requests with custom AI-based call center software

.NET

Business process automation (BPA)

Desktop

Information technology

NLP

Python

React

Telecom

Innowise has developed a custom AI-based call center software for a large telecom company, connecting operators and clients within a unified system.

Customer

Industry

Telecommunications

Region

Client since

2020

Our client is an internet provider, delivering a wide range of internet services to customers located within the European Union.

Detailed information about the client cannot be disclosed under the provisions of the NDA.

Challenge: Automating call center operations with custom contact center software

Our client, a European internet service provider (ISP), was facing the need to streamline its customer support operations and enhance the overall customer experience. To achieve this, they decided to outsource the development of custom call center scheduling software. The desired solution should enable the ISP’s operators to access client information and service details while being integrated with the current CRM and task tracking system. Additionally, the system was supposed to empower operators to create automated tasks for resolving customer issues as quickly as possible.

The old system version was plagued by poorly written code that made it difficult to maintain and update the system. The visual design was outdated, impeding operators’ ability to navigate and use the system easily. Furthermore, its monolithic and complex architecture made it difficult to scale and adapt to meet changing demands. The inflexible system also posed challenges in integrating with the current infrastructure. Therefore, the client reached out to Innowise to create a custom call center software solution based on the following requirements:

Automated call routing. Our client aimed to maximize call processing automation through the utilization of artificial intelligence.
Seamless integration with internal infrastructure. This integration is crucial for retrieving up-to-date client data, service history, and account information during customer interactions.
Card-based information retrieval. The system should present automatically generated cards that display relevant client information to operators during calls.
Task automation. Operators should be able to create and assign tasks to specific employees or departments within the organization directly from the contact center software.
Our client aimed to maximize call processing automation through the utilization of artificial intelligence.
Security and data privacy. The call center software must comply with the security policies and adhere to the General Data Protection Regulation (GDPR).
Intuitive interface and training. The software was required to offer an interface that would minimize the learning curve for operators. It should be easy to use and provide clear guidance on navigating through customer information and creating tasks.

Solution: AI-driven customer service interactions and automated request processing with call center automation software

Innowise has developed a robust AI-based call center automation software that processes calls from clients, connecting them with the appropriate call center operators after AI-automated voice recognition and request identification. We utilized a .Net-based microservices architecture to implement the back-end components, including SIP and soft PBX servers. The front-end part was developed using React and Electron, and is presented as a convenient and fast responsive desktop application. The system's primary objective is to receive and route calls through service-related numbers. The calls are then forwarded to the system's back-end and distributed to the call queue based on predefined regulations. The GDPR-compliant solution comprises various interconnected components, featuring core functionalities to facilitate fast response and communication with operators.

AI-powered voice recognition and call routing

Our team implemented artificial intelligence for contact centers to automate the client support operations by implementing smart voice recognition, interactive voice response, and call routing capabilities. Through automation, artificial intelligence now takes the lead in interacting with callers. The message processor AI bot was built based on modern natural language processing techniques, using the concepts of BERT models, and various speech recognition tools. Upon dialing, AI engages in a conversation, poses relevant questions, and recognizes the caller’s voice and request. Through artificial intelligence, the call center automates operations such as changing tariffs, ordering additional traffic, disabling services, and other simple tasks. This automation significantly streamlines operations, while for more complex inquiries, the AI seamlessly routes the call to the appropriate operator or department, ensuring customers receive specialized assistance.

This intelligent call routing ensures that callers are swiftly connected to the most qualified personnel who can efficiently address their specific needs. As a result, our client’s operators now serve as the final touchpoint in the customer’s journey, focusing on delivering exceptional service and resolving complex issues while fully automating simple requests.

Client request cards

The customer contact center software we delivered provides call operators with an automated card creation functionality. It enables employees to swiftly access and update relevant client details. When an operator receives a call, the system instantly identifies the client by phone number and their associated card, presenting it prominently on the operator’s interface. Each card within the software interface represents a specific customer interaction and displays pertinent information, such as personal details, service plans, billing history, and previous interactions, reducing manual data entry.

The card also includes an action plan that the AI system displays for the operator based on voice recognition and the nature of the request. This card includes an action plan tailored to the specific request, empowering the operator with step-by-step guidance to address the customer’s needs. For example, if the request involves solving network problems, the card might outline troubleshooting steps or reference relevant resources for the operator to resolve the issue effectively.

Automated task setting and progress tracking

In addition to card-based information retrieval, the custom call center software solution ensures seamless synchronization with the company’s CRM and task tracking system, enabling efficient progress tracking. This integration empowers operators to create and assign tasks to specific employees or departments directly from the desktop application in just a few clicks.

When an operator identifies a customer’s request that requires further action or assistance beyond the initial interaction, they can easily generate a task via a single click. The operator selects the relevant card associated with the customer interaction and initiates the assignment process. The system automatically transfers all pertinent details from the card, including the customer’s request, preferred resolution, and any specific individuals or departments responsible for handling the task.

As the assigned tasks progress, operators can update the task status within the system, which is then synchronized with the current CRM and task tracking systems. The software provides a user-friendly interface that allows operators to easily indicate the resolution stage for each task. This status-based approach provides supervisors and managers with an at-a-glance overview of the progress made on specific requests.

The ability to track task progress within the unified system helps supervisors and managers monitor the overall process, identify bottlenecks, and take necessary actions to ensure timely resolutions. It also provides valuable insights into workload distribution, allowing for efficient resource allocation and workload balancing across teams or departments.

Call recordings and statistics

The call center automation software securely stores all calls received by the center. All call recordings are sorted by date and stored in scalable data centers, and a synchronization protocol is in place, ensuring that data from the card can be accessed by any service, regardless of the display format used by different departments. This feature is designed to quickly resolve issues thanks to quick information access for various departments and enhance the training for call operators, with all calls logged and recorded for reference and learning purposes.

In addition to recording calls, the system offers customizable statistics and analytics tools. Using these instruments, the client can display reports and analyze the performance of their call center employees, track the number of interactions processed by AI, received and missed by each employee and optimize the team’s performance.

Technologies

Front-end

React, Typescript, Redux, Electron, SignalR, jssip, bootstrap

Back-end

.Net Core, C#, .Net WebAPI, NHibernate, FluentMigrations, SignalR, RabbitMQ, MassTransit, Kamailio, Asterisk, XUnit

Databases

PostgreSQL, Redis

Platforms

Desktop

Operating systems

Windows, Linux

Revision control systems

Git

Server Monitoring

Prometheus, Grafana

Log management

ELK

Continuous deployment

Gitlab Pipelines

Containerization and container management

Docker, k8s

AI & ML

Python, Tensorflow, Keras, Pytorch, CMUSphix,gTTS, Librosa, NLTK, Gensim, Pandas, Numpy, Flask

Process

Following a thorough review of the client’s requirements, Innowise chose Agile as the project management approach, which allowed our team to quickly adapt to changing requirements and facilitate collaboration among many team members. The project tracking was handled through Jira. The client was kept updated on the project’s progress via Google Chat and Telegram.

The project was divided into four phases: planning, development, testing, and deployment. We worked closely with the client during the planning phase to define the project scope and strict requirements. In the development phase, our team focused on building the system’s core components, including the back-end and front-end parts. The team tested the call center software rigorously to ensure its reliability. Finally, in the deployment phase, the team set up the system on the client’s servers and provided support during the initial launch.

Team

Back-End Developers

Front-End Developers

UI/UX Designer

QA Engineers

Project Manager

Business Analysts

Solutions Architect

DevOps Engineer

VoIP Engineer

Full-Stack Developers

ML Engineers

Results: 41% faster request resolution and 27% of automatically processed tasks with artificial intelligence call center solution

Our custom call center software solution has significantly improved the overall customer service process, with 27% of requests being processed automatically. As a result, the workload required to process queries has significantly decreased, enabling staff members to focus on tackling more intricate and challenging customer issues. The artificial intelligence call center solution enhances the overall efficiency and effectiveness of the ISP’s customer support operations by seamlessly synchronizing with the company CRM and task tracking systems, enabling task creation and assignment, and providing progress-tracking capabilities. It streamlines task management, helps optimize employee workload using statistics, and enables proactive monitoring and resolution of customer requests, leading to improved customer satisfaction and operational performance.Furthermore, it has also helped to improve coordination and collaboration between the company’s departments, leading to a 41% reduction in response time overlaps. The platform’s user-friendly interface and intuitive design have improved the efficiency of customer support. Operators can respond more accurately and quickly, and new operators can quickly become proficient with the system due to its reduced learning curve.With its customizable nature, the platform can be easily adapted to different call centers and departments, beyond just telecom services.

Project duration

January 2020 – February 2021

41%

faster request resolution

27%

automatically processed requests

Related cases

Contact us!

Book a call or fill out the form below and we’ll get back to you once we’ve processed your request.

What happens next?

Having received and processed your request, we will get back to you shortly to detail your project needs and sign an NDA to ensure the confidentiality of information.

After examining requirements, our analysts and developers devise a project proposal with the scope of works, team size, time, and cost estimates.

We arrange a meeting with you to discuss the offer and come to an agreement.

We sign a contract and start working on your project as quickly as possible.

27% automatically processed customer requests with custom AI-based call center software

Customer

Challenge: Automating call center operations with custom contact center software

Solution: AI-driven customer service interactions and automated request processing with call center automation software