Marketers who want to optimise their content and understand their audiences are watching the evolution of large-scale image analysis intently. The potential value of having unsolicited access to the spontaneous lives of consumers is clear, the how is less so.

At the risk of stating the obvious, images are of fundamental importance to marketers in this highly visual world. Much of what people post on their sites is their most aspirational selves- the people they want to be and the moments that want to have and be seen having. All of this presents a unique opportunity for creatives to be part of a specific conversation that their consumers are already having -a window into how to sell people the right stuff, in the right kind of way.

The problem, however, is the images themselves- there are just too many of them, updating in a constantly refreshing cycle. Take all the data and images that are stored within image-friendly platforms like Twitter and Instagram. There are more than 3 billion images shared on these social media sites every single day. The work required to find insights at this unprecedented scale, sifting through millions of images, image captions and videos- is stretching human capacity and taking precious time away from the creative process.

This scale of content means that creative departments now have a tough balancing act on their hands, trying to marry the creative side of the process– forming powerful narratives– with the infinitely less sexy tech side of data analytics. The two areas do not always sit well together. According to Dotted’s 2018 Marketing Survey, almost three-quarters- that’s 72%- of marketers feel that a culture of data collection and analysis is killing their creativity. Many blamed inefficient and costly data-collection processes.

AI solutions like Jaggu are here to help. Machine Learning technology cantake the mundane out of marketing -helping them gain insights into their audiences quickly, at scale and in real time. These insights become an invaluable tool for understanding how brands’ (or their competitors’) products are being interpreted by consumers- the language used, the visual associations being made, the locations chosen, the timings of posts and the emotions that underpin them. Jaggu technology can also help brands find out who their consumers actually are, how they feel, what they care about, and how they are relating to and influencing one another. It can even help marketers predict how their consumers might react to future situations.

Suddenly creatives have the key to understanding how to market to their specific audiences in the right way, bridging the gap between what their brand needs to say and what their consumers actually want to hear.

Take for example our recent collaboration with a global skincare brand. They wanted to find information that would help them sell millennials skincare products. In just a few days, we were able to hand them data analysis that confirmed that over half of teens are influenced by ethics when buying a brand. Our technology also deduced that young people associate skincare with products that prep their faces for make-up — social-media-friendly masks that get them glow-y enough for the next selfie. By making these marketing insights immediate for the brand, this data suddenly became the inspiration for and not confirmation of their marketing strategy- propelling their creative designs in the right direction from the outset and saving them valuable time.

As you can see our AI technology is in no way a replacement for creative marketers, it remains a creative’s job to turn this information into meaningful creative strategy. With our help, it will no longer feel impossible to look at endless images but instead it will feel like a lucrative opportunity to better connect with consumers.

If marketers invest in an automated analyst they’ll be one giant leap ahead of their competitors.

Intelligent video analytics is a huge step forward in how we can better use software as a way to make sense of the world around us.

The world around us is constantly shifting and transforming and is full of so much more information than a static, single frame can give us. If businesses truly want to understand why, what or how something is happening, they need to analyse our world in motion.

Of course, a computer recognising a polar bear or a cat from a still picture is still relatively impressive. The bigger test however lies in AI’s ability to understand that cat in a shifting, real-time video stream- noting it as not only “a cat” but also as “animal”, “mammal” and “feline”. This sophisticated level of understanding requires a really considerable amount of computing power.

Enter Jaggu Video — an AI platform which gives you the power to automatically understand what is happening in live videos. It is the digital eyes with the brains to match. Our technology identifies not only different objects in video but activities, patterns of movement, behaviour, locations, anomalies, demography, logos, genders and colours. It filters and processes all this data in real time- automatically interpreting an unfolding visual scene.

Action Training

The science behind this technology is known as deep learning. It involves training simulated neural networks to recognise different ‘stuff’ from a large number of examples. A team of paid staff watch videos and add keywords- providing the system with the examples it needs to learn. Here at Jaggu this semantic understanding is highly complex- our AI has a hierarchical knowledge and identifies categories as well as objects in live, action-packed footage, first noting “vehicle” then recognising it as “a car” and then finally naming it an “Aston Martin”.

Shelf Life

Retail is a competitive sector driven by constantly shifting shopper behaviour. It’s therefore more important than ever that retailers understand what is taking place in their shops and react quickly –customer and in-store insight being the new vital tools for a successful competitive advantage. In fact, Peak’s data science team recently found that retailers using AI grow 30% faster than those who don’t.

Jaggu Video monitors and measures in-store staff and consumer behaviour providing valuable insight into where retail organisations can improve their marketing, operational and staffing strategies. By giving retail business owners a live, constant stream of ‘shelf truth’ this technology frees up time spent on manual shelf inspections, improves product availability (and reduction of Out of Stocks) and can even reorder stock for you automatically. By analysing this situational and behaviour data, Jaggu helps companies understand the motivation for why, when and how a consumer makes a purchase decision at any given time. And, of course, this information is entirely objective- free of human error or bias.

Data Traffic

A live stream of automated video analysis for the council

Jaggu Video can also apply to traffic management- helping to measure pedestrian flow and vehicle traffic to dynamically improve congestion and reduce pollution. Jaggu’s AI-based systems utilises intersection cameras to manage traffic flow, by providing insights for various parts of the day and night. Furthermore, with the Chancellor today announcing a £420 million fund in the Budget to fix a record pothole problem, technology that could improve incident reporting and facilitate preventive maintenance of the road infrastructure seems more relevant than ever.

A city could also gain real-time, meaningful data for optimising the organisation of planned infrastructure. For example, this might involve planning where to position street lights in accordance with where crowds tend to congregate for events. When combined with other applications like smart parking, we hope our platform could also help with shaping infrastructure changes or future development.

AI Insight. Human Interpretation.

While our AI is great at identifying what’s going on in a video at a high level (e.g. someone is brushing their teeth or looking at their phone), we do not claim to be able to extract all the context needed to make those uniquely human calls. Ours is a collaborative technology that aids, not makes, decisions.

For example, Jaggu Video can identify that a “person is running,” but it will not be able to interpret if their pace is due to missing the bus or being late for work. This vital next step is still left up to people that see the data- we just help you get to an actionable response faster.

Our Intelligent video analysis is better focusing human attention on what matters most, to dynamically change the way users make sense of the world around them. After all, actions speak louder than words and we want to help you listen.

‘Jaggu’ is a discovery analytics company that comprehends complex data in real-time. This technology offers brands and agencies tangible results from which to base creative decision-making. A ten-strong team of engineers, researchers, executives and creatives offer digital data solutions to creative problems to some of the biggest names in marketing.

Jagadeesh Gorla began his educational journey at a remote state school in Peddapur, India. 10 years, a few English lessons and a MS and PHD later and Jagadeesh now lives and works in London, heading up his very own AI start-up specialising in automation and discovery analytics.

“Yes, my story does deserve a book series” Jagadeesh jokes, “perhaps in the ‘Inspirational non-fiction’ section”.

Here are a few of his founding thoughts…

Jaggu, what are you trying to solve ?

In today’s digital world, we produce, comprehend and disseminate information around-the-clock at work, at home and on the move. You probably respond to emails, take a picture and comment on some social media all before you’ve even had your breakfast. A substantial part of the knowledge economy is about comprehending all this constantly evolving information so that creative-problem-solving and decision making is fully informed. In all honesty this process of extracting information and making subtle judgements is mundane, expensive and time consuming. Using machine learning in these early creative stages relieves so much of that unnecessary resource pressure and frees up the human workforce to concentrate on the much more exciting issue of development.

What does assistive technology look like in day-to-day working life?

Over the last two decades, machines have subtly entered into our lives as assistants, meeting little resistance along the way. Take the simple example of search engines- they assist us daily by providing an ordered list of the ‘most likely’ contents to satisfy our needs (based on their comprehension of what those needs actually are). Indeed, there are very few people in the world today that could imagine their lives without the assistance of their trusty ‘Google’ right-hand man.

At present, there are an estimated 300+ million, and growing,‘knowledge workers’ in the world- people whose jobs rely on the organisation and comprehension of data. These repetitive, mundane tasks can potentially be automated by machines instead. If you are using Outlook or Gmail at work, you may not realise it, but you are actually already being assisted by machines… so why stop there? ‘Knowledge Industry’ workers (E.g, Marketing insight) are currently underserved by intelligent software. This is due to the unique and multi-layered nature of the tasks they perform, which has hitherto been deemed too complex for machine learning.

What is the purpose of ‘Jaggu’?

“Software (machines) as the ultimate, ameniable colleague”

Every business has its own particular set of procedures and solutions they use to solve their customers problems. Automation of these procedures, through machines, would help to accelerate growth and free up time for further development. It would also enable operational efficiency, speed of innovation and a shift in focus to creative problem solving. An intelligent automation software would therefore act as a never tiring, ever-present assistant! Machines don’t need a human level of intelligence and/or emotion to help grow the business- in fact it is precisely their lack of it that makes them such valuable asset. By 2025, it is estimated that the Automation of Knowledge Work will have a $5–$7 trillion impact on the world economy.

I have been fortunate enough to be part of the recent revolution in machine learning. Here at ‘Jaggu’ we have first-hand experience in building complex intelligent systems currently being used by 200+ million customers across three different industries (Gaming, Media and Retail). Using our collective experience at Jaggu and our customers’ feedback, we are determined to build intelligent software to automate repetitive and mundane data work- so you don’t have to. May the journey continue…

Parting thoughts…

We are always on a lookout for ambitious researchers and engineers to experience our exciting journey. A taste of what it would be like working with us can be described by the slight modification of old British saying (learning the ropes) … You don’t learn the ropes, you build them!