Simply enter your keyword and we will help you find what you need.

What are you looking for?

Good things happen when you narrow your focus
Welcome to Conference

Write us on info@evently.com

Follow Us

AI Testing

← Back to timetable

MAIN TRACK TALK

Ensuring Accuracy and Quality in Multilingual Call Transcription and Summarization

In the digital communication age, transcription and summarisation have gained unprecedented importance. Historically, manual methods often led to inaccuracies and delays. However, with the rise of AI and Machine Learning, this landscape began its transformation. Enrique’s company, a pioneer in virtual telephony, embarked on a unique journey by introducing multilingual call transcription in English, Spanish, and French. This ambitious venture underscored the pressing need for impeccable accuracy and quality. Enrique’s QA team, leveraging OpenAPI, Python, and Cypress, crafted innovative strategies to validate this feature, addressing a global business challenge.

In the realm of virtual telephony, Enrique’s team introduced a feature to transcribe calls in English, Spanish, and French, further providing users with summaries and key topics. Faced with the challenge of validating this process, Enrique (QA Lead) and Kevin (QA Specialist in Linguistics) developed an innovative testing strategy. Utilising OpenAPI, Python, and Cypress, they designed tests focusing on various scenarios to assure the quality of this groundbreaking feature. Join them as they share insights and methodologies, setting new standards in the virtual telephony domain.


What you’ll learn


From this talk you will learn how:

Gained a deep understanding of the complexities and nuances associated with multilingual call transcription and summarization in a virtual telephony environment.


Innovative methodologies, spearheaded by Enrique and Kevin, to validate the accuracy and efficiency of transcriptions, summaries, and key topic generation.


Practical knowledge on how to leverage OpenAPI, Python, and Cypress for generating synthetic data and orchestrating comprehensive tests for similar applications.


Strategies and insights to address challenges posed by multilingual and multifaceted data.


Session details

Track 1

12:00h - 12:45h · May 30th

40 min talk + 5 min Q&A

Topic: AI Testing

General Level

Talk in English, Translated to Spanish

enrique_sanchez

Enrique Sánchez-Bayuela

Enrique is an Engineering and QA Manager for over 13 years, leading dynamic teams at companies like Aircall and Cabify. Currently, he specializes in AI testing using ChatGPT, innovating in simulating production traffic and evaluating audio transcriptions. Enrique also shares knowledge as a university professor. What he cherishes about testing is the challenge of ensuring product excellence and the continuous evolution in methodologies, especially with the advent of AI. Being an active contributor to the Spanish QA community, Enrique’s shared insights at various conferences.

kevin_perkins

Kevin Perkins

Kevin Perkins, originally from New Jersey, USA, has been residing in Madrid since 2012. He spent two years pursuing a dual Masters degree in Applied French Linguistics at Complutense University in Madrid and Sorbonne in Paris, at the same time he taught English. Upon completing his degree, Kevin worked as linguist, where he refined his skills for several years. In 2019, he transitioned to the dynamic field of artificial intelligence, assuming a quality assurance role for a virtual assistant. Since 2021, Kevin has played a crucial role at Aircall, overseeing the quality assurance of the company’s AI services.