At Omnilytics, our technology is our most valuable asset. We process millions worth of data points daily to provide our clients with intelligent and accurate insights. At the centre of our tech development lies Dr Mohammad Mahboubian, the Chief Technology Officer of Omnilytics.
With a PhD in computer science, Mahboubian joined Omnilytics as a data architect to help innovate and expand Omnilytics’ data mining capabilities and was soon promoted to CTO in 2019. Today, he leads the entire tech department, including the three main functions, data science, data mining and data engineering.
As the CTO, he sets the overall vision and direction of our technology development and ensures a consumer-focused outlook in delivering our product to the market. On top of his product-related responsibilities, Mahboubian also develops internal tech integrations between the various departments within the company.
To introduce Omnilytics’ tech and the team behind it, we recently sat with Dr Mohammad Mahboubian and discussed the relationship between retail and data, along with his vision for 2021 and the future.
How Does Omnilytics’ Technology Work?
The basis of our technology can be broken down into three steps – data collection, data processing and data visualisation. “The insights surfaced on the Omnilytics dashboard are obtained from over 100,000 brands and retailers around the world, including major names such as ASOS, Nordstrom and Farfetch,” says Mahboubian.
“Each SKU consists of multiple data points, and every day we capture millions of these SKUs to extract valuable retail insights” he continues. These data points are collected by an in-house scraping engine, designed and developed by our very own data mining team.
A unique feature of the Omnilytics dashboard is the depth of data available. The data visualisation displayed on the dashboard is based on a set of AI algorithms and state-of-the-art machine learning, in which text and image models are trained to recognise the distinct attributes of fashion products such as ‘dress’, ‘A-line’ or ‘party’ – also known as taxonomy.
Due to these highly subjective attributes that can differ from brand to brand, Mahboubian states one of the biggest challenges of developing the dashboard is having to continuously train and improve the accuracy of the AI-based systems to process a large volume of data on a daily basis. This is vital in ensuring the quality of data and user-experience for each of our clients.
More on Omnilytics: Meet the rest of our team
How Will Data Analytics and AI Impact the Future of Retail?
“This year was an extremely challenging time for the retail world and our clients included, so we had to configure and upgrade our product to meet their needs for the current conditions”, says Dr Mohammad Mahboubian.
With the retail industry turned on its head due to diminishing demand, lockdowns and weak economies, brands and retailers from various industries have learned to embrace data analytics. “Many retailers were facing uncharted territories for the first time as historical data no longer held its merit,” he mentions.
Prior to the Covid-19 pandemic, historical data would be heavily used in many retail processes such as sales forecasting and buying, but today’s market fluctuations make it impossible to rely on these historical patterns alone. As a result, more companies started seeking data solutions to identify and validate consumer demand through real-time data.
“We can definitely see a change in people’s attitudes towards data and AI in general. There used to be a fear of AI replacing jobs, but now they’ve realised the technology cannot work independently,” Mahboubian explains. “In fact, much of the machine learning is based on humans performing these activities and the focus of the system remains extremely narrow.”
The current capabilities of AI are still in its early stages, humans not only play an important role in training these models but also analysing the results, making them the final decision maker. Ultimately, the main benefit of using these solutions is to help reduce cost and wastage by only producing products that are in demand and in the right quantities.
“Our mission is to help brands and retailers to validate and empower their decision making by providing in-depth market information,” Mahboubian says. “This extends to seasonal trade performances, pricing strategies and even trend spotting.”
He further explains, “hopefully with our data, our clients can increase the chances of a product selling out and get closer to the commercial goals of their companies.”
Looking Ahead to 2021
“We had to learn a lot and move quickly due to the pandemic,” he mentions. “We intend to take these learnings to further develop our product and better serve our clients in navigating the aftermath of Covid-19.”
On next year’s plans, “we will be incorporating more markets, on top of the 40 countries we are currently observing, as well as developing new features for the Omnilytics dashboard,” says Mahboubian. Among the projects he will oversee next year include the new data API integration to help expand Omnilytics’ usability across platforms.
“As a tech company, it’s important for us to deliver the best possible product, this mentality keeps us striving for innovation and to remain a pioneer in our field,” said Mahboubian.
About the Author
Sufiana Sharuddin
Sufiana Sharuddin is a published fashion writer, honing her interest in the industry during her time at Condé Nast College of Fashion and Design. She currently covers a variety of topics within the industry including business, technology, trends and current affairs.
