top of page

Q&A with Imane Bouayad, Data Scientist @ Cadi Ayyad University

Updated: Feb 24, 2022

Exclusive Trusted Magazine Q&A with Imane Bouayad, Data Scientist @ Cadi Ayyad University


What are the latest technologies with impact in the AI landscape?


AI and its applications are transforming services across different fields, with recent advances in the fields of natural language processing and finance having the most impact.


In Natural language processing, large language models are considered the building block of various AI technologies as they’re used in, but not restricted to:


  • Intelligent keyboards, email response suggestion and spelling autocorrection.

  • Extractive or abstractive summarization and generation of texts.

  • Machine translation systems.

  • Speech and handwriting recognition.

  • Conversational agents (chatbots).

  • Recommender systems.


The Generative Pre-Trained Transformer 3 (GPT-3) by OpenAI is a widely used language model, it can perform a variety of tasks like answering complex questions, generate fiction, creative texts and computer code.


During the pandemic, financial services grew considerably as more people were turning to online services. This transition led to a surge in the development and use of mobile and AI based applications in the field of finance and banking.


Consequently, it’s expected that AI based financial services will multiply as the consumer demand is constantly growing. AI technology in finance is used to leverage varying financial services:


  • Capturing trends, forecasting and predictions.

  • Financial virtual assistants.

  • Investment management and guidance.

  • Fraud and cyberattacks prevention.


With the help of AI and machine learning in finance, it will allow for government officials, companies and business managers to plan future activities and make better decisions.



How is AI changing some organizations and processes in a disruptive way?


Advances in AI and its applications are disrupting organizations and processes in different ways, on one hand questions about ethical AI are constantly raised as various AI applications are used and deployed with no ethical supervision.


Let’s take language models as an example, it’s undeniable that bias in these models is continuously created and amplified as they’re being used in many applications, which results in a disproportionate weight in favor of or against specific ideas and communities.


On the other hand, the legal field is adopting AI to assist with decision making, legal search and analysis and legal document management. In this regard, AI is also changing how legal services are managed and delivered by automating processes and remapping existing business models. This adoption will allow for more efficiency, transparency and ease of access to legal services.


AI technology is a double-edged sword, it can facilitate interdisciplinary prospects and benefits as it can disturb existing processes and create contentions.



What are your recommendations to succeed in AI projects?


Bird’s eye view: having a clear vision of the project is usually an important yet overlooked step in designing AI applications, this step helps with choosing the right design or strategy for your applications and enables accomplishment of short- and long-term goals.


Effective communication: communication sits at the center of success in any project, but for AI projects where you have an interdisciplinary team, it’s even more essential. As a data scientist, it’s important to understand the business side of any project, to learn and use business metrics in showcasing improvement at the beginning, during and the end of the project across time, costs and challenges, and finally ask for feedback.


Implementation: deploying and maintaining AI application is a core function in all AI projects, in doing so, it’s essential to opt for an MLOps approach that’s based on the following steps:


  • Record and store different types of data using the cloud.

  • Creating and managing data pipelines.

  • Data and models validation.

  • Continuously Deploying and maintaining ML models.

  • Evaluating and finetuning.


Practicing and using MLOps allows for better implementation of AI applications, automating AI processes and delivering value to various fields.

206 views0 comments
bottom of page