Success Stories

Automated sentiment analysis
and voice classification using AI technologies for an improved IVR system

Introduction:

In today’s digital world, companies are increasingly looking for innovative ways to differentiate themselves from competitors, optimize operational processes and meet growing customer demands. One critical aspect is optimizing interactive voice response (IVR) systems to provide a more personalized and efficient user experience. This case study deals with the implementation of an automated sentiment analysis and voice classification system with the aim of optimizing an IVR system.

Company profile:

A leading telecommunications company operates a call center with a high call volume and diverse customer inquiries. The company’s goal is to make the existing IVR system more responsive through optimization and to better meet individual requirements to improve customer service.

The company’s challenge:

• Limited personalization: The existing IVR system was unable to adapt to customer moods and preferences.
• Manual call routing: Human intervention was required to effectively route calls, leading to delays and potential customer dissatisfaction.
• Inefficient sound processing: The system struggled to differentiate between human voices, IVR announcements and background noise, which negatively impacted the overall user experience.

Objectives:

• Sentiment analysis: Implement sentiment analysis to understand and categorize the emotional tone of the customer voice during interactions.
• Voice classification: Develop a robust model to distinguish between human voices, IVR-generated announcements, and background noise.
• Automation: Integration of the sentiment analysis and voice classification modules to automate the IVR system to enable real-time adjustments based on customer sentiment.

Implemented solution:

1. Signal processing:

• Feature extraction: Using signal processing techniques to extract relevant features from the audio signal, including pitch, tempo, and spectral characteristics.
• Speech pause detection (VAD): Implementation of VAD algorithms to identify speech segments to enable targeted sentiment analysis.

2. Sentiment analysis:

• Text-to-Speech (TTS): Integration of TTS systems to convert audio to text for sentiment analysis.
• Natural Language Processing (NLP): Implementation of NLP technologies for sentiment analysis of the transcribed text to gain insights into the customer’s emotional state.

3. Voice classification:

• Feature extraction for voice type: Developing features to distinguish between human voices, IVR-generated announcements, and background noise.
• Machine learning model: Training a machine learning classifier using a large dataset to accurately classify different types of voices.

4. Integration and automation:

• System integration: Integration of the sentiment analysis and voice classification modules into the existing IVR system.
• Automation rules: Definition of rules for automatic call routing, personalized responses and adjustments to system behavior based on detected sentiment and type of voice and concern.

Technology stack:

Python libraries/packages used:

Numpy, Pandas, Librosa, Matplotlib, Seaborn, Pydub, Keras, TensorFlow, Scikit-learn, Asyncio, Websockets, Streamlit, Pyaudio.

Results:

• Improved customer experience: Automated sentiment analysis enables the IVR system to adapt to the customer’s emotional context, providing a more personalized experience.
• Efficient call routing: Automated voice classification ensures accurate call routing without human intervention, resulting in faster response times and higher customer satisfaction.
• Improved sound processing: The system can now differentiate between background noise and human voices to prevent interruptions and create an improved customer experience.

Conclusion:

The successful implementation of automated sentiment analysis and voice classification in the IVR system has significantly improved the customer experience. By leveraging advanced technologies, the company has achieved greater personalization, efficiency and adaptability in its customer service processes.

This case study illustrates the transformative power of integrating innovative solutions to overcome real-world challenges and improve customer interactions in the digital age.

Flexibility and expertise:
Remote Java developers from Alfa IT outsourcing as the key to success

Introduction:

A company in the field of data acquisition from IT sensors was planning the further development of a self-developed system and needed additional developer resources with various specialist skills.

The company’s challenge:

Our client was unable to fill the vacant IT positions themselves as there were not enough qualified applicants. Although the company had already opened up to remote work and extensive recruitment measures meant that no suitable IT specialists could be found. There was also no internal capacity to drive forward the further development of the system.

Objectives:

• Staffing with qualified IT experts
• Go-live with new system on budget and on schedule

Solution:

For various internal projects and to support the development of the in-house system, the client required a full-time experienced Java Developer for a long-term collaboration. The selection process, including CV screening, video interviews and test assignment, was completed within 14 days of the client’s request, allowing the Java developer to start work with our client at short notice.

During the collaboration, the client had an additional need for support in the front-end area for the development of their own system. Within a few days of the customer inquiry, we provided our customer with a full-time frontend developer with extensive knowledge of Vue.JS, in line with the requirements profile, for a predefined period of around 3 months until the release date.

The frontend developer’s initial task was to implement and integrate the existing design.

Implementation:

• The IT coordinator we provided ensured the seamless integration of the referred developers into the existing IT team on the client side
• Close cooperation between the placed frontend developer and Java backend developer as well as the internal IT team on the customer side, as the areas and tasks were interdependent
• After the go-live, the front-end developer was reduced to part-time (2 days per week) after 3 months in accordance with the agreement to complete subsequent adjustments and integrations

Result:

The enhanced system was put into operation on time and on budget on the planned release date.

Testimonial:

“Normally, we always take a conservative approach to filling our IT vacancies, as we have had bad experiences with service providers in this area in the past. However, due to budget and time constraints, as well as the fact that we simply could not find qualified IT specialists ourselves, we decided to give an alternative approach another chance: the Alfa IT Outsourcing concept.

We have not regretted this decision and it is considered one of the most important steps of the year for us. Our expectations were completely exceeded. In particular, the flexibility we gained, the way Alfa IT-Outsourcing responded to our requirements, the quality of the developers’ work and the added value provided by Alfa IT-Outsourcing’s IT coordinator convinced us across the board. With Alfa IT-Outsourcing’s customer-oriented concept, we were able to eliminate a lot of our worries and challenges.”

Conclusion and outlook:

The remote developers we provided were seamlessly integrated into the client’s existing IT team with the support of our IT coordinator. The Java developer will continue to be employed full-time after the end of the project to work on other internal IT projects.

During the collaboration, the client had a short-term need for additional Java development and asked for another remote Java developer to join the existing IT team for a period of 3 months. The client immediately received CVs of suitable candidates from us. After video interviews and test tasks, the Java developer selected by the client started work on the agreed start date.

The client is currently working successfully with two remote Java developers in collaboration with an IT coordinator from Alfa IT Outsourcing. The onboarding of the additional Java developer and the general cooperation between the customer’s IT staff and the booked Java developers are running efficiently and smoothly. As a result, the frequency of daily meetings with the free mentality bridge (IT coordinator) was reduced to two days a week.

Legacy Modernization:
Creating figures, data, facts with Business Intelligence solutions

Introduction:

A leading market research company was faced with the challenge of improving its outdated, self-developed Business Intelligence (BI) tool. After decades of use, the tool developed in Java/Scala had reached its limits. In order to be prepared for the future and to be able to offer a broader range of valid, meaningful data, reports and benchmarks, the company decided to modernize the tool. Neither the expertise for legacy modernization nor the necessary documentation and IT resources were available. The customer asked itself: Where and how should we start?

The company’s challenge:

The customer’s BI tool had significant limitations after a long period of use. The lack of documentation and the lack of expertise in the area of software modernization made it difficult to bring the tool up to date. In addition, the customer’s internal IT department was not sufficiently staffed to carry out this extensive modernization independently. The costs for implementing the project also had to remain manageable and within budget.

Project objectives:

• Future-proof: Develop a stable and user-friendly BI tool.
• Data availability: Enable the creation of an unlimited amount of well-founded, meaningful data, reports and benchmarks.
• Cost control: Development and implementation should remain within a manageable budget.
• Documentation: Creation of meaningful and complete documentation.
• Implementation approach: Create the basis for the successful implementation of software modernization.

Implemented solution and realization:

Originally, the client only wanted to hire a Java/Scala developer to support the internal IT team and make improvements to the existing tool. However, after further consultations, a comprehensive modernization of the BI tool was recommended in order to fully achieve the project goals.

Implementation steps:

1. Tool analysis with the help of a business analyst:

• A business analyst was employed to fully understand the functions and implications of the customer tool.
• Analysis of the current status of the software and creation of complete documentation of all modules and dependencies.

2. Defining the system architecture with the help of a system architect:

• A system architect was involved to visualize the future tool landscape.
• Recommendation of the optimal tech stack (.NET and React) after weighing up all the pros and cons.
• Definition of the number and functions of the required IT resources.

3. Project implementation with the help of an IT coordinator:

• Use of agile methods and introduction and use of a project management board.
• Close cooperation and continuous exchange between customer, IT specialists and IT coordinator.

4. Personal meetings:

• On-site visits to the customer to discuss what has been achieved so far and details of the next steps.
• Strengthening the partnership through personal exchange.

A business analyst and later a system architect were initially deployed to implement the project. The customer also benefited from the experience of the IT coordinator, who was provided as a service by Alfa IT Outsourcing at no additional cost.

Thanks to the use of agile methods, a project management board, best practices and close communication between the customer, IT specialists and IT coordinator, all project participants and team members always had a complete overview of the project status, progress and results. The personal exchange on site further strengthened the already good relationship of trust with the customer.

Result:

• Comprehensive documentation: The business analyst successfully documented the current status to the customer’s complete satisfaction.
• Precise recommendations: The system architect formulated precise information and recommendations, which were visualized using flowcharts, tables and calculations.
• Solid basis for decision-making: The customer can now use this information to make well-founded decisions on how to proceed and implement the software modernization.

Testimonial:

“Initially, we planned to hire a Java/Scala developer to address existing issues and make incremental small improvements to our BI tool, which quickly proved unfavourable due to the lack of documentation and uncertainty about dependencies and potential consequences. Through extensive research, we came across Alfa IT Outsourcing. We received intensive and competent advice from the very first meeting. We were immediately convinced by the advantages of comprehensive software modernization and the strategic approach. We are happy to have found an experienced modernization partner in Alfa IT, as well as a professional team that really took the time to understand our requirements. This has certainly saved us a lot of headaches.”

Conclusion and outlook:

The decision to modernize our BI tool proved to be without a doubt the right approach for this complex project, which was also confirmed several times by the customer. He was impressed by the expertise of the IT specialists, especially given the initial uncertainty about how to approach software modernization in the first place. Both the business analyst and the system architect quickly got to grips with the BI tool and the workflows and developed a deep understanding. The system architect’s ideas and recommendations were particularly praised.

Thanks to this successful collaboration, Alfa IT Outsourcing was not only able to reach a decisive milestone in the customer’s project implementation, but also to strengthen their trust to establish itself as a competent partner for business intelligence solutions and software modernization.