How Wazzat Labs is adding ‘eyes’ to your online search
What do you do when you notice a ‘must-have’ dress or a purse at a party and feel like owning one?
You have to browse through online websites for hours to find it. This is primarily because it is very difficult to describe the dress you are looking for in words. With the advent of smartphones, the last decade has seen an explosion in the mobile internet ecosystem. However, the ways in which users search for content online are still primarily text-based and fairly inefficient.
What if you could take a photo of the dress you like and your phone or tablet could automatically show you all the dresses available online that look like it? Wazzat Labs, a Hyderabad-based image recognition startup, can help e-commerce websites do that for you. The company has launched visual tools that can be easily integrated into existing websites and apps to improve the online search and recommendation results by replacing the existing text-based search with more intuitive visual search. Simply put, Wazzat Labs is an image recognition company that adds ‘eyes’ to your apps and websites.
The genesis of Wazzat
While completing his Masters in Computer Science at IIIT-H, Jay came up with the idea of a mobile-friendly platform which can solve some of the commonly faced problems today. His visual search solutions were even recognized at international conferences. Later, his batchmate Naveen joined him to strengthen the architecture to handle large databases and provide seamless integration into websites and apps. Both met Mauktik Kulkarni through a common friend. Mauktik says,
Since real-time, efficient and visually intuitive search is the biggest unmet need and the biggest business opportunity in the fashion e-commerce industry, we all decided to focus on that first. Our solutions can be seamlessly integrated into any e-commerce platform and is at the heart of developing omnichannel strategies. However, the core technology lends itself to solving a wide range of problems in travel, enterprise mobility, defense, mobile healthcare and other sectors. The fashion industry has recently seen a lot of startup activity and a few success stories. We are glad to be making steady and serious progress in this fast-paced segment.
What goes on behind the scenes?
Wazzat Labs is combining its cutting-edge in-house image recognition research and the growing processing power of computers and handheld devices to solve the problem of visual search. Mauktik, COO of Wazzat Labs, says,
Through several years of image recognition research at IIIT-Hyderabad and Wazzat Labs, we can demonstrate that our visual fashion search results are significantly better than when a user types in a description of the item. It is extremely fast, doesn’t take up a lot of space on the phone and requires minimal bandwidth.
Wazzat has developed a mobile-friendly platform that enables apps to ‘see’ things, recognize them rapidly and make instantaneous decisions based on it.
Jay Guru Panda, CEO of Wazzat Labs, says,
The platform is highly optimized for mobile devices; it can learn to recognize up to 100,000 images, condense the knowledge to a few megabytes and download it to the handheld device. The device can then instantaneously recognize these images without even requiring internet connectivity. With this technology, the time it takes for mobile searches goes down from 5-10 seconds to less than a second.
Incorporated in February 2014, the company was one of the top three companies at the MIT-GSL accelerator program and has joined the Target Accelerator Program. Mauktik feel that this is great validation of their strategy and the team. Wazzat Labs is running pilots with Target.com, one of the largest retailers in the US. They are also in discussions with several other fashion outlets and looking forward to launch pilots with them soon.
Mauktik claims that recently launched, Wazzat Fashion, will be a suite of solutions for the exponentially growing fashion e-tail industry.
What’s your take on Wazzat Lab’s visual search tool? Let us know in your comments below!
Website: Wazzat Labs