Pavlo Karalupov

A web application, created from scratch for Ukrainian medical reform, that allows patients to make appointments with doctors from home and monitor clinic activity.

I developed this application using HTML, CSS, JavaScript (AngularJS), TypeScript, Microsoft SQL, and C# (ASP.NET and Entity Framework). With this app, I won several hackathons and it now has over 2.5 million active users.

Amazon Transcribe makes it easy for developers to add speech-to-text capabilities to their applications. Audio data is virtually impossible for computers to search and analyze. Therefore, recorded speech needs to be converted to text before it can be used in applications. Historically, customers had to work with transcription providers that required them to sign expensive contracts and were hard to integrate into their technology stacks to accomplish this task. Many of these providers use outdated technology that does not adapt well to different scenarios, such as low-fidelity phone audio that's common in contact centers, and this results in poor accuracy.

Amazon Transcribe uses a deep learning process, called automatic speech recognition (ASR), to convert speech to text quickly and accurately. Amazon Transcribe can be used to transcribe customer service calls, automate subtitling, and generate metadata for media assets to create a fully searchable archive.

Voice control allows you to operate the Swisscom TV-Box with your voice. Swiss people usually speak more than one language fluently and, if needed, can switch easily between them—even in the same sentence. This phenomenon is called code-switching in the field of speech recognition.

The same phenomenon can be seen in the Swisscom TV Voice Assistant. In that scenario, the user might switch between languages when searching for a foreign actor; for example, "Find me a movie with Gérard Depardieu," where the intent is spoken in English and the entity is spoken in French.

After investigating different approaches for code-switching and multilingual speech recognition, I improved the system's word-error rate by 10% in code-switching scenarios.

PayPong is the progressive, online, instant financial services platform for those who are active and enjoy life. It offers a quick and convenient solution for financial needs, such as:
• Money transfers from card to card and by bank details.
• Payment for telephone, internet, utility, and many other services.
• Getting an online loan in 15 minutes, credited directly to your bank card.
• Factoring and leasing.

A financial web application for browsing and assessing the risk of buying a particular stock.

I identified my customer's needs and requirements, selected the tech stack, implemented the system from scratch, wrote unit and integration tests, and configured the testing and deployment pipeline.

Ticket sales applications for the Latin American region that used several payment providers as well as several payment methods (cash, card, partner seller, and more).

I was responsible for application enhancement on the back-end side such as the customer reservation profile to perform various actions with tickets, integration with more providers, ticket check-in, creating the recurring back-end tasks with the dashboard, among others.