The product-focused CTO is now more common than not. What industry they find themselves in, determines how they spend their time and what products they develop.
Ed Bishop — co-founder and CTO at Tessian — for example, spends the majority of his time focusing on what machine learning modules should be built into the products that Tessian offers. He does this in order to find out which modules will have the biggest impact, and which modules shouldn’t be built in the first place.
Being a CTO for an artificial intelligence solutions provider
As part of Information Age’s Artificial Intelligence Month, we speak to Harald Gölles, CTO of AI provider, omni:us, about his role in the organisation. Read here
The results of Bishop and his team’s efforts can be see in three machine intelligent email filters:
Guardian: This predicts misdirected emails, so sensitive data is not inadvertently sent to the wrong recipient.
Enforcer: This classifies emails as authorised or unauthorised (or to personal/malicious accounts). In a world when more and more people use freemail domains for legitimate business, a machine intelligent classifier is required.
Defender: This predicts inbound emails to prevent against strong-form impersonation spear phishing attacks.
Cutting through the noise in AI
As a CTO in the AI space, succeeding in his role relies on cutting through the noise or hype surrounding AI, and “making the right technology bets,” according to Bishop.
CTOs on our Doorstep: Croud CTO on turning AI hype to reality
The ongoing developments surrounding automation in the digital marketing industry are mirrored by the innovations being delivered by AI and machine learning, according to Dave Lowe — CTO — Croud. Read here
There is also a need to “scale an engineering team in cutting edge technology stacks (big data processing and AI), and design scalable systems capable of deploying AI solutions (and corresponding data) to the largest companies in the world,” he continues.
Applying machine learning to products
Not many companies are using applied machine learning in their products. Most applications use machine learning to enhance an existing business model or feature; for example recommendation systems for e-commerce sites.
“Our product is machine learning. With this comes certain challenges – for example our enterprise clients need our predictions to be explainable to the end user, our machine learning can no longer be a ‘black box’,” says Bishop.
“This is challenging for our data science team, as it limits the types of algorithms they can use.”
CTOs in this space will need to deeply understand the fundamentals (and limitations) of AI and machine learning, according to Bishop. And, most CTOs have yet to be fully exposed to these technologies.