A few words with Shivang Desai and Chandralika Hazarika, founders of Bigthinx
Shivang Desai (SD) is Bigthinx CEO/CTO. Indian. 34 years old. Studied MBA (Finance & Strategy) at Case Western Reserve University, Ohio, USA. Shivang has previously founded three startups: one of the first drone manufacturers in India, a marketing and branding company for medical professionals (exited) in USA, and an influencer marketing platform based on the stock markets. Apart from management experience, he is a coder, technical artist and 3D modeler. He has also authored a fantasy fiction novel “Halo of Ashes”. His hobbies include extreme sports, writing, and reading.
Chandralika Hazarika (CH) is Bigthinx COO. Indian. 34 years old. Studied MBA (Marketing & HR) at IISWBM, Kolkata, India. Chandralika previously worked at a Singapore based MNC for eight years as their Brand Head. She helped them expand to various countries from Singapore, managing several diverse teams globally. In addition, she also owns her own fashion label, Velvet Piano. With an obsession with textiles and fabrics, her hobbies include fashion designing, research and trend forecasting in fashion.
TFR: What is Bigthinx?
SD/CH: Bigthinx is an AI company operating in the fashion and retail industry, offering 3D mobile body scanning, clothing size predictions, virtual avatars, digital clothing, and virtual fashion shows and photo-shoots.
TFR: What is the entrepreneur story that inspired the launch of the company?
SD: Chandralika and I, Shivang, both are entrepreneurs at heart and can’t help but find solutions to problems we personally encounter. When shopping online, we realized just how much wastage and losses businesses were incurring because we couldn’t decide what clothing would work best for us to purchase. We ourselves returned a lot of products we bought online because the information conveyed to us was either incorrect or inaccurate.
This set the stage for our entry into the fashion industry and we sought to devise ways of tackling these problems we faced ourselves, and millions of other shoppers face as well when shopping online.
E-commerce returns cost only the US economy $309 Billion in lost sales in 2019. This figure globally is worth hundreds of billions of dollars more. 30%-50% of all apparel items bought online are returned which end up going to discount stores or are destroyed. Additionally, 84% of all clothing returns are destroyed through land-filling or incineration. This has a huge impact on the environment.
When we realized how big of a problem this was, we decided to build a product that could target it. We started working in fashion in 2017 and started using AI to achieve our objectives.
TFR: Steve Laughlin, general manager of IBM Global Consumer Industries explained on Forbes that “AI can assist design teams by enhancing and reducing overall lead times, and expand their creative discovery by analyzing and remembering insights from thousands of images and videos using computer vision. Also, designers can detect how they can integrate trending colors, key patterns, and styles”.
How is AI evolving and how fashion designers remain artistic when machine learning algorithms are used?
SD/CH: Machine learning and AI are extremely useful tools to determine the what, why, where and how of fashion designing. However, from a creative standpoint, designers still need to envision and create their designs based on their own sensibilities and their brand image. Machine learning should be used to draw usable insights into trends and then the designers need to ideate products based on the learnings that it provides.
Just like any other tool, these tools are meant to allow fashion designers to do their work faster, better, and with more data to draw upon. However, the artistic part of fashion design still remains highly individualized and that is unlikely to ever be completely replaced by machine learning algorithms.
TFR: On your website, you mention “We are digitizing the human body with data on anatomy, shape, and clothing size, fit and drape, that enables brands to personalize shopping, envision designs, optimize supply chains and save costs”.
If I´m right, your solutions benefits designers, buyers, planners, and in consequence, customers. Could you explain how the roles mentioned are improving their activities?
SD/CH: We are transforming fashion retail with not only digital clothing and virtual avatars but more importantly, body data. Such data is very useful to make big changes in the entire supply chain. With just two smartphone photos, our sizing technology gives you 44 precise body measurements that can be linked to any brand and help them reduce returns.
Taken further, this data enables brands to modify their designs and product specifications to suit their customers better, decide on manufacturing volumes to produce only as many products as will actually sell, reduce wastage and unsold clothing, and improve their bottom line. Brands can also trial designs with customers to pre-sell and receive feedback before manufacturing so that they know which designs are likely to sell more.
TFR: What is Lyflike?
SD/CH: Bigthinx’s Lyflike AI creates a personalized virtual 3D avatar from just 3 pictures with one’s mobile phone – a front body pic, a side pic and a selfie – and automatically recreates digital 3D clothing from any 2D image instantly to show how it would look, fit and drape as if in reality. These avatars can be used to visualize clothing look and fit, to address visualization and fit requirements for customers purchasing clothing online, and even in-store via touchscreens and gestures.
TFR: Some of the trends described in The Fashion Retailer are Fashion as a Service (rental, reuse and customization) and Sustainability. Both trends are enhanced by customization (made-to-order), 3D Knitting and AI.
Your solution Lyfsize gets 3D body scans and 40 precise body measurements and ratios from just two smartphone pictures. In previous posts I described AI success factors and mentioned startups like Zozo. For example, Zozo suit measurement system uses algorithms to capture precise sizes and make custom-fit clothing by using 3D scan too.
What is your competitive advantage or key differentiator (e.g. better algorithm, machine learning, scanning accuracy, UI, UX…)?
SD/CH: Unlike the Zozo suit which required a special bodysuit to be worn for capturing body measurements, Lyfsize does not require any special preparations. The user just needs to use their mobile phone and take two pictures – a frontal and a side shot of their bodies – wearing well fitted clothing such as gymwear. The software calculates and processes over 44 body measurements in less that one second. These measurements can be used for custom made clothing and to find out one’s size and fit in any clothing brand. In the rentals space, this allows customers to rent from other people who match their own body types so that they know what they are renting will fit them well.
The Lyfsize software uses a combination of computer vision, machine learning and deep learning to provide 44 accurate body measurements from just two pictures through the use of superior technology. The UI/UX is also designed to make it extremely easy to take a body scan through two pictures alone, and a user can easily complete this process within 5 seconds.
While several companies are solving different challenges in the fashion industry, what sets us apart is our advanced technology. Our tools are well suited to moving the industry towards a much more sustainable future.
TFR: Bigthinx recently powered a virtual fashion show at the FASHINNOVATION Worldwide Talks. There are many advantages (e.g. speed and agility, cost-reduction) and disadvantages (e.g. clothing and materials “need” to be seen and touched; fashion shows are social events).
What is your opinion about the future of fashion shows and its digital alternative?
SD/CH: We believe that even though digital fashion shows will not completely replace real ones, they will certainly become popular alternatives for brands that want to showcase numerous collections but don’t want to take on the burden and expense of real fashion shows. They will also become a leveling channel for e-commerce brands and upcoming designers who cannot afford to put on traditional fashion shows and can benefit from a digital medium.
Virtual shows do not require any physical interactions whatsoever between designer, models, stylists, spectators, and buyers. As a result, they can be produced and showcased with all interactions happening purely virtually and thus are free of restrictions on movement and hygiene fears.
This approach is also far cheaper and much more sustainable than traditional methods, and we expect to see widespread adoption in the near future.
Various formats are being explored such as models dressing in designer outfits and recording themselves on cameras in their own homes, ramp walks in front of green screens with the backgrounds being replaced with digital imagery, and of course, purely digital options such as ours. We also have an option to place 3D clothing on a 2D actual model in a real world environment.
Fashion weeks have always been extravagant affairs, often more about people seeing and being seen than just the conducting of business. In a COVID-free world, there is going to persist a stigma towards close personal interactions and mass gatherings, and a call for more sustainable options. We expect that fashion weeks will follow the recent example set by Gucci to go seasonless and showcase fewer shows each year. While fashion shows will not disappear by any means, they will see a far greater proclivity towards sustainability and reducing wastefulness, a part of which includes supplementing their normal formats with purely digital or hybrid mediums. Fashion will be much more inclusive and sustainable.
Social interaction can be well carried out using digital channels as has been evidenced with the pandemic. Hybrid mediums also allow users to interact with fabric swatches shipped to their homes while virtually experiencing how these fabrics can look and move through a virtual fashion show, thus effectively providing workable alternatives to real fashion shows.
TFR: Another important restriction in regards AI is that many retailers have low maturity levels of master data management. If a product is not well described, even from the PLM (Product Lifecycle Management system) perspective, it would be difficult to have the right insights. In this sense, many retailers are far from implementing machine learning due to poor data quality management.
How do you help those companies or what would you suggest to them?
SD/CH: Design houses and fashion brands are looking at pushing the boundaries of what is possible by using digital mediums and technology to recreate anything they can dream up.
Among these requirements is to have environments and backgrounds that are engaging, interactive, and potentially personalized to offer user experiences tailored to the viewer’s moods or personality.
The primary challenge we faced was in educating designers and models on how and when to deliver pictures. In normal fashion shows, designers are often working until the very last moment that the model walks on to the runway to make sure the look is exactly what they envision. However, in order to produce a virtual show, pictures and details of the clothing need to be provided in advance to allow time to digitize them, which is something that designers are not used to. We even displayed garments in the virtual show which were from only concept sketches and even a prototype had not been made yet.
We encourage designers to invest in skills for 3D modelling, pattern making, and CAD softwares. Since digital avenues are starting to become an integral part of the fashion industry, it will be vital for designers to have these skills to stay abreast of the rapid progress being made with technology, and its application in purely digital or hybrid mediums. We are always there to guide in the transition. In Fact we have been approached by fashion colleges globally to guide the students in their transition to a completely digital future of fashion.
From an AI/ML standpoint, we help retailers with ways to organize their data so that it can be correctly and quickly interpreted, resulting in measurable benefits once they are ready to start using AI in their products.
This opens up a new way for 3D design experts familiar with computer software to find themselves jobs in the fashion industry. This is also the ideal time for designers to upskill their organizations by hiring 3D designers and AI experts or investing in acquiring these skills themselves.
TFR: India is a high-quality tech talent hub. What helped to rise this IT service cluster and specializing talent in India (e.g. government, schools, companies…)?
SD/CH: India has always had a proclivity towards engineering and software development. Many big companies globally employ Indians in these areas due to the skillsets developed by such individuals as a way to upskill themselves. Traditionally, this has been the best option for promising individuals to be hired by large global corporations.
As a result, India has a fantastic talent pool of techies who can rise to different challenges and make the best use of the opportunities anywhere in the world.
Although as a company, Bigthinx is based in Europe, a big part of our technology team will continue to be based in India. However, we also intend to have a R&D team based in Europe and Canada.
With remote working now becoming the norm, this also provides a huge opportunity for Indians to work from their homes so that they can maximize their earnings as well as benefit from a wonderful personal life close to their loved ones.
TFR: Big data analyst, specialist in machine learning, developers…are very demanding in the big data era. How do you manage to find these profiles? How do you keep them motivated as Millennials aren´t afraid to change jobs?
SD/CH: We hire carefully and with great discretion. AI is suddenly a buzzword and there are plenty of people trying to sport the tag of Big Data and Machine Learning professionals without having the necessary skill-sets.
The difference between the larger tech pool and the candidates we hire is that we look out for individuals who are truly excited to be working on new technologies, and have demonstrated growth in the knowledge through personal pursuits. This is also motivating for them as they are the ones excited about being the first to develop and utilize the most cutting edge technology. We provide the resources they need in order to maximize their potential so the candidates we hire always have challenges and opportunities to grow, which is more fulfilling than changing jobs for just the immediate income they are earning.
TFR: As a start-up, what do you think are the key success factors that brings you to compete to bigger companies?
CH: We invest in technology for a competitive advantage. This is what allows us to do what many larger companies just aren’t able to because of our agile approaches. Our team is extremely lean but yet manages to produce some of the best products for the fashion industry because they are always challenged to push the boundaries and make things that are better, faster and more efficient.
A lot of our time is spent on ideation and brainstorming, dreaming up out of the box solutions and better ways of doing things.
Us founders are both first architects, then management people with a finance, strategy and marketing background. So although creativity is very important to us, we never neglect to analyze the business feasibility and market fit. This approach differentiates us.
Shivang is also a serial tech entrepreneur, this being his fourth start-up. Previously he had successful exits in USA and India, and brings years of field experience to the table.
TFR: Do you feel Fashion is becoming more science than art?
SD/CH: Fashion companies have traditionally been slow to adopt digital technology. However, with the COVID-19 pandemic, they have been forced to re-think their approaches to customer experience as well as designing, manufacturing, and supply chain management.
There is a huge scope for improvement in the way fashion companies operate such as using body data to improve clothing specifications to fit their shoppers better, identifying how many pieces of each size to manufacture to prevent wastage and unsold merchandise, digitally sample clothing designs pre-production, and even trial designs using virtual avatars of real customers to get a better idea of which products will sell better.
While the creativity of fashion design still remains art in its essence, its delivery can definitely become more scientific and benefit from technology.
Technology like ours can evolve to being a standard part of e-commerce shopping experiences, branch out to in-store applications to allow quick and easy clothing trials, and be used to own purely digital collections for use in virtual communications and interactions. With the coming of holograms in the near future, phygital interactions will also become commonplace. Moreover, the advantages of this technology lie in improving the entire supply chain of the fashion industry. Human body data can be used by brands to tweak their clothing specifications and sizes to fit their shoppers better, produce only sizes that they know will sell and thus reduce wastage, and pre-trial designs with customers so that they manufacture only the products which have received positive responses.
Bigthinx is also expanding its product range this year with some niche offerings but details will only be released upon launch. As always, in fashion-tech Bigthinx is committed to helping as many brands, designers, and students, irrespective of scale or budget. We believe in creating a sustainable, circular fashion economy that is significantly more inclusive and diverse than the one we have today.
This period of Covid-19 uncertainty can become an opportunity for brands that are losing in-store footfalls and sales to begin selling products directly to consumers online. Without sacrificing any of the advantages of in-person shopping, consumers can see how products look and fit on them, and find out exactly which size to buy from the convenience of their own homes. Payments can take place online, with deliveries easily arranged from the nearest physical stores or warehouses. This maintains the consumer experience while also opening new, powerful channels of selling.
Bigthinx has also ventured into digital models and virtual photo shoots, enabling brands to shoot catalogs and organize virtual runway shows completely digitally and at a fraction of the cost of real shoots.
See a few awards and achievements of Bigthinx’s below.
● Finalist of the Luxury Fashion Category at The Mark Challenge, Monaco. Awards to be announced in June. Only company from Asia.
● Selected as one of the top 10 fashion startups globally in association with Startupbootcamp Fashion tech Milan.
● Awarded for the best use of AI in fashion by Awards.ai London in its 4th global annual AI awards 2019-20 ceremony.
● Won silver at the Amazon AI Awards 2019 in the Retail and Ecommerce category, 2019
● Selected as one of only 10 companies from around the world by Taovation China to participate in the “Changemakers for China 2019” program.
● Chosen as a top-30 startup from India to participate in Slush, a Finnish flagship Startup event, bringing together global tech leaders to Helsinki, Finland.
● Selected as one of the top 10 startups in India to compete in the Startup World Cup 2019 by Pegasus Tech Ventures at the World AI & RPA show (India’s Largest A.I Gathering).
● Selected as one of the 20 top tech startups in India to join the India-Sweden Startup flight organized by invest Stockholm.
● Honored by InterCon as one of the 50 top tech companies for the year 2019.