# Discovering Business Value in AI Solutions with Richmond Alake
Written on
Chapter 1: The Importance of a Business Value Mindset
In the realm of AI development, adopting a business value mindset is paramount. In a recent podcast featuring Lewis Tunstall, co-author of "NLP with Transformers," we delved into this crucial topic. It's essential to understand the significant difference between casual projects and those that have real-world applications. Solutions that employ advanced deep learning techniques may excel in theory but will only find relevance in practical scenarios if they deliver tangible business value. Recognizing this early in your career empowers you to direct your energy towards initiatives that truly matter to organizations with actual customers.
As an example, I've been developing an exciting tool that leverages Transformers and various AI technologies. Instead of merely showcasing the impressive capabilities of transformers or detailing the dataset creation process for training deep learning models, I've shifted my focus. I now concentrate on the business value that potential users of this demo tool seek from AI solutions.
Consider a sentiment analysis algorithm that evaluates customer feedback on specific products, enabling marketing or sales teams to act upon consumer opinions. If a product garners a significant number of negative reviews, the logical response for a company would be to reduce stock orders and address the underlying issues that led to the negative feedback.
Traditionally, my presentations on AI solutions would emphasize the technical specifications and architecture of transformers, along with the training methods for the models I use. However, current and prospective business clients often prioritize their customers' needs above intricate technical details. As professionals, we should also keep our clients' customers at the forefront of our focus.
Chapter 2: The Fascination of Transformers
Transformers are not only captivating but their multimodal capabilities are even more compelling. While they are predominantly utilized for natural language processing (NLP) tasks, there has been a notable shift towards applying transformers in vision-related challenges. During my discussion with Lewis Tunstall, we explored the intricacies of transformers and Lewis's journey in the machine learning landscape.
Section 2.1: Recommended Reading for Data Engineering
For those eager to deepen their understanding of Data Engineering, I recommend the following titles:
- Data Pipelines with Apache Airflow by Julian de Ruiter and Bas Harenslak
- Fundamentals of Data Engineering by Joe Reis and Matt Housley (Early Release)
These resources will enhance your technical, architectural, and foundational knowledge, which is vital for aspiring data engineers. Stay tuned for more articles on data engineering from me.
Thank you for taking the time to read this article. I trust you found the insights valuable!