Ocado’s AI: Under the hood with Gabriel Straub, Chief Data Officer

Ocado Group’s pioneering AI journey spans the course of nearly 25 years when we first set out to innovate the world of online grocery. Through smart warehouse automation and advanced machine learning, we are transforming the landscape of traditional grocery fulfilment. And this is just the beginning of our AI story.  

Earlier this year Ocado’s Chief Data Officer Gabriel Straub, won the AI Expert of the Year award from the Digital Leaders organisation. Fresh off his win, we sat down with Gabriel to get under the hood of Ocado’s AI journey, exploring our innovations, our biggest wins and what’s still to come. 

Gabriel, your bio is pretty expansive – from advising boards and start-ups to leading large data science teams. Can you give us summary of your career to date? 

Yes of course! I’m a little bit of a chameleon.

My passion is helping organisations realise the importance of their data. I’ve got over 10 years of experience leading data science teams, working as Head of Data Science at the BBC, Data Director at notonthehighstreet.com, and Head of Data Science at Tesco. I studied Mathematics and Business, receiving an MA in Mathematics from Cambridge and an MBA from London Business School. I'm also a guest lecturer at London Business School and an independent commissioner at the UK’s Geospatial Commission.

What was it that attracted you to Ocado Group?

I joined Ocado in 2020 to deliver data and AI innovations. Data and AI are at the heart of Ocado’s platform and decision-making, which made the role an easy-sell for me. 

Primarily, I was intrigued by Ocado’s approach to navigating the complex challenges of online grocery. Typically, grocers trying to grow online have focussed mainly on the front of the store to deepen their profits – looking into assortment, promotions, pricing and loyalty. For most this has had limited success in growing customers, and almost no success in improving the underlying economics of the channel. Ocado focussed on developing custom AI solutions to solve the complexities of online grocery. 

Today we’ve got a bespoke end-to-end tech platform, where AI is integrated at every step of the grocery journey from the very beginning – on the webshop – to the very end, at delivery. We work with leading retail and logistics businesses that use this technology every day. The value they gain from this integrated system is evident in their operations, delivering higher efficiencies and building hyper-personalised customer experiences.  

I’m responsible for driving the company’s data strategy, equipping the data science team with the right tools, processes and culture to succeed, as well as overseeing Ocado’s cutting-edge use cases of AI.

‍Can you tell us more about how you’re deploying AI in unique and innovative ways across Ocado Group?

At Ocado Group, we’re innovators at our core. We had to build our own bespoke solutions rather than rely on off-the-shelf technologies, as there was no end-to-end technology capable of handling the complexities of the online grocery supply chain.   

We’ve created an end-to-end platform that includes AI-driven stock forecasting, automated customer fulfilment centres, delivery optimisation and personalised shopping experiences. Now organisations have the opportunity to do everything - from their ecommerce to their fulfilment to their last mile - from one solution.

The search functionality on our ecommerce solution is a great example of how our teams are using AI to build better customer experiences. Customers visiting our retailers’ webshops are now presented with a range of relevant results when searching for items. This helps them to find what they need quicker and more conveniently.

When it comes to efficiency, there are so many AI-powered elements of our platform we can pinpoint but my favourites are our robots which pick and pack groceries. They work as a team, moving at speeds of up to 4m/s to transport containers of grocery products across an electronic grid, enabling our partners to pick a 50-item orders in just 5 minutes.

How does Ocado's AI give the business a competitive edge?

At Ocado, AI underpins our customer offering, helping us to deliver a wide range and unparalleled service while enabling retailers to make a sustainable profit.

With over 3,000 patents to our technology granted and filed and 2,500+ technologists in our team, we are operating at the cutting edge of software and hardware engineering.

All of these innovations are having an impact efficiency and the bottom line: 

• AI powers our webshop - driving revenue by encouraging customers to repeat purchases, find new favourites and expand their baskets through intuitive offers and promotions. 
• AI maximises the efficiency of our fulfilment centres - enabling partners to pick 50-item orders in just five minutes, saving time and money. 
• AI makes the supply chain more accurate - allowing retailers to predict demand, keep purge low and improve product availability.
• Machine learning and advanced data science optimise the last mile - helping our partners to deliver on time for customers and achieve efficiencies at scale.‍

What are your predictions for AI in the future? How can businesses go beyond the hype of Gen AI?

I regularly discuss with my team how important it is to not oversimplify our approach to AI. To remain competitive for the future, diversification is key. Here are my 3 predictions around the future of AI: 

1. Traditional machine learning methods will remain dominant:

So much of what we see today is businesses swept up in the hype of Generative AI. While it will remain popular in the coming years, traditional machine learning techniques will remain the tried-and-true tools for most use cases. These techniques have already been tested rigorously, providing reliable results time and time again. They’re also incredibly versatile, capable of being applied to a range of circumstances. Most importantly, these methods are simple to understand and implement, often requiring less power and data compared to the complex models used in generative AI.

2. The systems and processes we know today will be completely overhauled: 

We’re in the very early stages of generative AI. A stage I like to call ‘early electricity’. In the early days of electricity we saw limited efficiency gains as factories were still designed around the constraints of older technology. It wasn’t until new structures were put in place that we really started to see the benefits. Today, Gen AI is similar as it’s still being integrated into existing structures. If we want to reap the real benefits of this technology we’ll need to overhaul how we design our processes and systems, shifting our existing architecture to leverage entirely new AI capabilities.

3. Suppliers are in for a gold rush.

While these new forms of AI are still evolving and unfolding, its full impact remains uncertain.  In historical gold rushes, it was the suppliers of tools that profited more consistently than the miners themselves. We will see the same thing here with the suppliers winning out. If your company is providing essential tools and platforms like cloud services, AI training and specialised software, you’ll be poised to capture the most value. It’s all about enabling others to explore AI, helping companies to understand the applications will be most valuable to them. 

How is Ocado continuing to innovate, and what can we expect in the years to come?

Our AI-powered robotic arms are the next efficiency gamechanger, helping our partners to unlock unprecedented levels of growth. These robotic picking arms have developed rapidly over the last 2 years and are continuing to innovate beyond their current capabilities – scaling across industries and adapting to new products and environments.

I’m excited to see what comes next as we double down on the most advanced deep learning approaches available today to streamline our efforts and set us up well for future Machine Learning research.

‍

Our platform is unique is because it’s end-to-end – encompassing the grocery ecosystem. Data is integral from the front end to, supply chain, last mile and fulfilment operations. In the future we will continue to innovate around this area, using the data that flows across our platform to shape better experiences for retailers and the customers.