Don't Fear the Future: Gen AI is a Win for Software Development

Generative AI (Gen AI) is a hot topic, and 2024 is already shaping up to be a big year for its further development and implementation. Major companies like Apple, Amazon, and Qualcomm are recognizing this potential and actively working to advance its capabilities.

Just envision – advanced algorithms capable of generating unique content, from unprecedented musical compositions and distinctive artworks to novel narratives. This is the future of generative artificial intelligence (Gen AI). Using extensive datasets, these sophisticated systems produce original content intelligibly tailored for human consumption.

And here's the good news: things are looking bright. When used responsibly, Gen AI can boost efficiency and slash errors by creating smarter, more innovative solutions. The key is using clean and accurate data.  

On the latest episode of Collider Convo, featuring hosts Jacob Smith and Rhyan Robison, they dive deep with Loïc Giraud, from Calibo, to explore everything about Gen AI and its impact on software development.

Not All Data is Created Equal – But Good Data Management is Priceless

In 2023, Giraud, an engineer with a decade of experience in the field, started developing Calibo, a self-service development platform. With clients across industries, they've mastered the product-market fit.

“When I look at AI, I really look at how you can augment human intelligence by using computing power,” says Giraud.

Though artificial intelligence isn’t new, Giraud notes two major developments in recent years have made AI exponentially more powerful.

“Number one, I think the amount of data that you can ingest and train your model on is astronomic and can help you have much more accuracy in terms of the output of your models,” says Giraud. “Number two is that we have a new generation of software models, which we call large language models, or generative AI, which has really helped to start to democratize AI.”

Though AI has transformative properties which can add greater efficiency to how we work and live, it’s critical that the data that AI ingests needs to be tidy.

“These models are being trained on data, and the data needs to be clean, and it needs to be governed,” says Giraud.

Fortunately for customers, that’s Calibo’s area of expertise. By simplifying data ingestion, transformation, and storage processes and enhancing data transparency, Calibo empowers organizations to confidently embrace AI.

“Calibo simplifies the way you’re ingesting, transforming, and storing your data,” says Giraud. “So instead of writing a lot of code, you can use configuration steps that prevent human errors and still create your Single Source of Truth.”

Calibo takes it one step further by bringing complete transparency to the data, including the data source, to give end-users confidence that they know where the information comes from. In addition, Calibo also provides a quality framework and features to its platform, which helps businesses understand the level of maturity of the data and its accessibility by the users who create and consume it.

This is where Calibo’s value stream mapping functionality comes into play.

“When you have an engineering team that is building a product, how do you know the effectiveness and efficiency of that engineering group?” asks Giraud. “Usually, it occurs via maturity assessments and questionnaires, so as you perform tasks and building and promoting codes, you’ll register a lot of data points.”

Giraud explains that value stream mapping extrapolates these data points to highlight the true value and maturity of the team developing the product. In addition, as engineers continue to create software or data products, value stream mapping can identify areas for improvement in the code.

"We also enhance our capability by bringing a quality framework and quality features to our platform, which helps the business understand the level of maturity of the data and the level of accessibility of the data by the users that create and consume it," describes Giraud.


Quelling Fears About a Future With Gen AI

For some engineers and even the public, generative AI may seem risky. If AI is not stirring up images of Arnold Schwarzenegger’s T-800, it may be stirring up worries surrounding job obsolescence. But Giraud’s not worried; he sees roles adapting to manage AI’s data models.

“If I really look at the history, I look at what happened with data and AI. You had the creation of data engineers, then data owners, and data scientists,” says Giraud. “I think the responsibility will definitely evolve, and it’s coming more from the angle of bringing more veracity to the information so that the output can be trusted by the organization that utilizes AI.”

Though there are those who still need convincing, for Giraud who has spent almost all of his career in data analytics, there’s no stopping the future.

Giraud recalls that 10 years ago, many people had the same hesitations about the cloud. 

“Ten years ago, when the cloud started to become more mass market, you saw a lot of people who were very reluctant to go to the cloud because of security reasons,” says Giraud. “If you look at it today, almost everything is in the cloud.”

Although it’s reasonable to have hesitations, for Giraud, the future includes generative AI.

“Gen AI is what the cloud was 10 years ago,” says Giraud. “It will take five or six years to mature, but it’s not something that we’re going to go backward on.”