The Anatomy of AI: Will It Unleash an Entrepreneurial Revolution?

[00:00:00] The Anatomy of AI Will it unleash an entrepreneurial revolution?

[00:00:08] Imagine a day in the life of an Indian farmer.

[00:00:11] He wakes up to his weather assistant informing him that it will rain in about two hours.

[00:00:18] Minutes later, another beep, his livestock sensor alerts him to a new virus on the block.

[00:00:24] As the day progresses, more beeps and alerts vie for his attention.

[00:00:29] It's no surprise for his farm data is on cloud for real-time advisory.

[00:00:35] His plot is mapped by smart drones, his crops are monitored by agribots and the harvesting

[00:00:40] is taken care of by GPS driven autonomous tractors.

[00:00:45] He is a farmer of the digital age, with data at his fingertips and technology in the palm

[00:00:52] of his hands.

[00:00:54] As much as this sounds like a chapter from a futuristic novel, it is in fact central

[00:01:00] to the government of India's vision of a new reality, the digital farmer.

[00:01:05] At its heart is artificial intelligence or AI or intelligent machines that enable high

[00:01:11] level cognitive processes like thinking, perceiving, learning, problem solving and decision-making

[00:01:19] thanks to advances in data and computer processing.

[00:01:23] The farmer appears in the strategy document of policy think tank Nithi Ayug, national

[00:01:29] strategy for artificial intelligence just as India launches its new drive hashtag AI

[00:01:36] for all.

[00:01:37] The message is loud and clear, it's time to embrace AI.

[00:01:43] The race heats up.

[00:01:45] Even if you are not aware of it, AI is involved in almost all aspects of your daily life, commuting,

[00:01:53] shopping online or even sending an email.

[00:01:56] Gmail has become the most popular email service in the world thanks to AI.

[00:02:02] From GPS, wayfinding to social networking and mobile banking from commanding voice assistance

[00:02:09] to clicking pictures on your mobile camera, from ride sharing to streaming entertainment,

[00:02:15] from self-driving cars to health tracking wearables, AI is all around you.

[00:02:21] It has also unleashed an entrepreneurial revolution with millions of bright young minds launching

[00:02:27] startups with very low costs and lofty ambitions in virtually every field of activity.

[00:02:33] They are identifying problems to solve, discovering user needs, gathering feedback, creating

[00:02:40] prototypes of their products and scaling up as fast as possible.

[00:02:45] The peak of the COVID-19 pandemic while detrimental to much of the world's progress, in fact

[00:02:51] accelerated the use of AI.

[00:02:53] The PWC 2022 AI Business Survey indicates that 52% of global organizations have accelerated

[00:03:01] their AI adoption initiatives.

[00:03:05] In India, the number of AI startups has grown from 471 in 2016 to 72,993 in 2022 according

[00:03:15] to the Ministry of Commerce.

[00:03:18] The Anatomy of AI.

[00:03:21] Despite the excitement AI has created, it is hard to define.

[00:03:25] So let's unbundle what it is first.

[00:03:27] AI is a thinking machine that can solve problems, learn and change.

[00:03:33] Second, problem solving is its key strength and the value it brings to the table.

[00:03:39] Third, its primary advantage is the ability to analyze massive quantities of data at a

[00:03:45] rate far beyond the capacity of the best of human brains.

[00:03:49] Fourth, it applies that knowledge to a vast battery of algorithms at blinding speed.

[00:03:55] The field, formally developed by MIT cognitive scientists in the 1950s, has caught the world's

[00:04:02] attention since the 1990s, especially after IBM's deep blue chest computer defeated Russian

[00:04:08] Grandmaster Gary Kasparov in 1997.

[00:04:12] The technology has progressed dramatically since then.

[00:04:15] AI is increasingly moving into realms once thought to be reserved for humans.

[00:04:22] For instance, machine learning mimics human cognition by learning from data to improve

[00:04:27] on tasks.

[00:04:29] Deep learning is where systems are taught to learn by example just the way humans do.

[00:04:35] Natural language processing happens when machines mimic the human ear and understand written

[00:04:41] or spoken language.

[00:04:43] Later vision is where machines are given thousands of images and analyze those pixel by pixel

[00:04:49] like the human eye.

[00:04:51] Neural networking is where systems replicate the way humans learn.

[00:04:56] Robotics process automation is where machines enact human actions efficiently to save time.

[00:05:03] And generative AI is where systems can mimic human efforts by predicting words and images.

[00:05:10] Unlocking a world of data.

[00:05:13] Generative AI has now given rise to conversational AI or technologies like chatbots or virtual

[00:05:21] agents that can talk to people and are largely used in customer service websites, online

[00:05:27] stores and social media channels.

[00:05:31] The release of chat GPT and AI model trained to learn what humans mean when they ask a question

[00:05:37] and provide complex long form answers has taken the tech world's breath away.

[00:05:43] Launched in November 2022 by OpenAI, the biggest nonprofit dedicated to friendly AI, chat GPT

[00:05:51] can write poetry, song lyrics, essays, TV episodes and more.

[00:05:57] According to a recent survey conducted on 1000 US business leaders by job advice platform

[00:06:04] resumebuilder.com nearly 50% have admitted to using chat GPT at their companies.

[00:06:11] And India is not lagging behind.

[00:06:14] In February 2023, velocity, a Bengaluru based financial technology firm that provides e-commerce

[00:06:20] founders with necessary business insights launched Lexi India's first AI assistant

[00:06:26] powered by chat GPT.

[00:06:29] A birup made acre CEO and co-founder of velocity writes in his blog post.

[00:06:35] Ever since chat GPT launched, our product teams have been brainstorming on how it can

[00:06:41] be leveraged to benefit our founders.

[00:06:45] Since velocity customers already use insights on a daily basis we integrated chat GPT

[00:06:51] with the same interface that they leverage for driving business spending revenues and

[00:06:56] so on.

[00:06:58] Now with chat GPT their customers will be able to use Lexi for answering their queries.

[00:07:05] The power of AI would not have been possible without data.

[00:07:09] Roughly data means information and never in human history has there been such an explosion

[00:07:14] of information as today.

[00:07:17] This information is being produced collated and made available first by the world of business

[00:07:23] and increasingly by governments and the civil society to see, sense, build better profit

[00:07:30] from and sometimes even to monitor citizens or consumers.

[00:07:34] The digital universe now generates 2.5 quintillion bytes every day according to software review

[00:07:41] platform financesonline.com

[00:07:45] In computing data happens to be the formation that has been translated into an efficient

[00:07:51] format.

[00:07:52] Put simply it is the raw material produced by abstracting the world into categories, measures

[00:07:58] and other representational forms like numbers, characters, symbols, images, sounds, electromagnetic

[00:08:05] waves, baits etc., that constitute the building blocks from which information and insight

[00:08:10] is created.

[00:08:12] Since AI can analyze massive amounts of data in minutes it is being harnessed for every

[00:08:18] service.

[00:08:19] According to the International Monetary Fund data should be considered a key input in

[00:08:24] modern economic production alongside land, capital labor and oil.

[00:08:29] From polls and elections, mergers and acquisitions and serving customers to preventing crimes everything

[00:08:36] is data centric today.

[00:08:38] Data has been the key in the world's pandemic response when thousands of scientists across

[00:08:44] the world joined hands with massive data sets to discover the genomic sequence of coronavirus,

[00:08:51] identify variants of concern, design new diagnostic tests and develop vaccines at

[00:08:56] breakneck speed.

[00:08:59] Data on tests cases, vaccines, beds, ventilators and deaths have informed decision making.

[00:09:05] No wonder data is being called many things from the new oil to the new gold and the new

[00:09:12] science.

[00:09:13] The fastest growing field.

[00:09:16] AI success is becoming the rule, not the exception.

[00:09:20] It is the fastest growing field in the tech world predicted to become a trillion dollar industry

[00:09:26] globally by 2035.

[00:09:29] This means that AI will be used by every industry.

[00:09:33] Businesses are using AI to predict data on customer behaviors that are seamless end-to-end

[00:09:39] and personalized.

[00:09:40] For remote workers and students to beef up online shopping and dining options, make

[00:09:46] customer call centers more efficient or to speed develop important new drugs.

[00:09:53] AI technologies are making it easier to search, manage, filter, classify and detect text,

[00:09:59] image, audio or video processing tasks.

[00:10:03] For instance, processing face recognition in public places.

[00:10:07] It is also being used to prevent fraudulent transactions and financial services.

[00:10:13] And it's not just business.

[00:10:14] AI is coming into the hands of communities at the coal phase of challenges.

[00:10:19] A number of innovative case studies across the world show how the use of AI technologies

[00:10:25] is bringing about exponential changes that are sustainable, speedy and scalable.

[00:10:32] In Paraguay, an AI-driven poverty measurement tool, Poverty Stoplight has helped nearly

[00:10:37] 24,000 families overcome intractable economic hardship.

[00:10:43] A UNICEF initiative is figuring out the troubling question, what keeps children out of school?

[00:10:50] In countries like Chad, Medagasca, Namibia, Nepal, Togo, Zambia, Kenya and Uganda with AI

[00:10:59] enabled school profile cards to help schools and parents fix problems in their education

[00:11:04] system.

[00:11:05] Similarly, indigenous communities of the Philippines, powered by data and AI tools, are engaging

[00:11:13] with mining companies on natural resource governance and demanding their share of royalty

[00:11:18] payments for the first time.

[00:11:21] The spotlight shines on India, the world's over performer in frontier technology readiness

[00:11:27] according to the UNCTAD technology and innovation report 2021.

[00:11:33] The AI market is steadily increasing, at $12.3 billion in 2022 and predicted to grow

[00:11:40] at 42% shows the AIN research report, the state of AI in India December 2022.

[00:11:49] It shows that IT companies make up 60% of the total AI market while non-IT sectors make

[00:11:56] up 23% of the total AI market share with banking, financial services and insurance or BFSI

[00:12:04] holding 28% among them.

[00:12:07] Where does AI go next?

[00:12:09] We stand on the brink of a technological revolution that will fundamentally alter the way we live,

[00:12:16] work and relate to one another.

[00:12:18] Rites Closed Shrop Founder and Executive Chairman of World Economic Forum, the International

[00:12:24] Organization for Public Private Cooperation in his book, The Fourth Industrial Revolution

[00:12:30] 2018.

[00:12:33] Shrop argues that we are now witnessing the beginning of the fourth industrial revolution.

[00:12:39] If the steam engine has driven the first industrial revolution, electricity the second and digital

[00:12:45] computing the third, AI is the force behind the fourth.

[00:12:49] In its scale, scope and complexity the transformation will be unlike anything humankind has experienced

[00:12:56] before.

[00:12:57] AI can already do many things more accurately than humans, so where does it go next?

[00:13:04] That is the question the real world AI warriors are asking as they delve deep into the power

[00:13:10] of cutting edge technologies to develop high performance and highly sophisticated designs

[00:13:15] that help solve complex challenges.

[00:13:18] They are mining data that has never been collected before, taking on the uphill job of building

[00:13:23] their own unique algorithms, quickly implementing those to create unique designs that have never

[00:13:30] existed before, gathering feedback and refining their models to perfection.

[00:13:36] For them, the areas that AI can impact are endless, and the potential to optimize and innovate

[00:13:42] is limitless.

[00:13:45] This chapter takes you on a non-technical roller coaster ride into the world of AI through

[00:13:50] the lens of outstanding young personalities who are bringing the technology to the forefront.

[00:13:56] They help enterprises that are making effective use of AI attracting venture capital interest

[00:14:02] and global attention.

[00:14:04] They are also exploring new ways of harnessing AI for the greater good, from assistive technology

[00:14:09] for people with disabilities to food robotics that can revolutionize the way we eat and

[00:14:15] cybersecurity solutions to safeguard and protect your business from malicious attacks.

[00:14:22] Part 1.

[00:14:23] Robot Cook Me Some Home Food

[00:14:26] Like most busy couples in the nation's high technology heartland Bengaluru, Yatinwarachia

[00:14:32] and Margi led tremendously overshed old lives.

[00:14:36] The former, a techie and the latter, an MBA working in healthcare.

[00:14:41] Both loved what they did, but what they did not like was the ever accelerating pace of

[00:14:46] their lives as well as their changing relationship with food.

[00:14:51] The act of cooking and eating once a pleasurable activity was fast becoming an irritating, albeit

[00:14:58] necessary function.

[00:15:00] They were too frazzled, too tired or too busy to cook nutritious meals for themselves.

[00:15:06] They were fed up with eating out and exasperated with the astronomical rates of hard help.

[00:15:12] They were also too alarmed by the scary statistics doing the rounds.

[00:15:16] How Bengaluru's IT sector was losing billions of dollars on account of poor health and poor

[00:15:22] diet?

[00:15:23] What was the way out?

[00:15:24] Yatin started exploring.

[00:15:27] He first shared his ordeal with his network of friends about a hundred strong, half of

[00:15:32] whom lived in the US.

[00:15:34] They all echoed the same sentiments, all seemed to be scrambling for the taste of home.

[00:15:40] Some had no access to authentic Indian food while others could not deal with the pain

[00:15:44] of cooking.

[00:15:46] Many complained about their unhealthy reliance on convenient, ready-to-eat, packaged processed

[00:15:51] or restaurant food that frequently took a toll on their digestion.

[00:15:57] What could Yatin do to solve a problem that seemed to be personal as well as collective?

[00:16:04] Home cooked food without lifting a finger.

[00:16:07] That was then.

[00:16:08] Today, he is on the way to solving the problem in a unique way.

[00:16:13] Every meal he and his wife take is nutritious and freshly prepared just like home food.

[00:16:20] But too, without having to lift a finger.

[00:16:23] We are eating out of Nosh, says Yatin.

[00:16:27] Nosh is the AI-based robotic chef that 37-year-old has created along with three friends for cooking

[00:16:34] a vast range of delicacies with tastes that bring back memories of the food they grew up

[00:16:40] with.

[00:16:41] I have stopped taking food to the office.

[00:16:43] Instead, I bring food from the office to home he laughs.

[00:16:48] A little larger than a microwave, Nosh can cook over 200 dishes and is a potential game

[00:16:55] changer.

[00:16:56] Not just for the varachias but also for anyone in a time crunch or one who does not know

[00:17:02] how to cook has no time to cook or does not wish to cook yet wants to savor the flavors

[00:17:08] that once simmered on their plate.

[00:17:12] Food has sentimental associations with happy or meaningful times.

[00:17:17] The comforting memory of our past and Nosh has the potential to change the way we live,

[00:17:22] work and eat.

[00:17:25] An inventor at heart.

[00:17:27] I am a techie at heart and Gujarati by senses and I love inventing new solutions to difficult

[00:17:33] problems that require multidisciplinary innovation.

[00:17:37] Right Siyatin about himself on his social media accounts.

[00:17:41] His family in Kantoa, a small village in Gujarat can vouch for that.

[00:17:46] Even as a child he would tinker with sticks, stones, blocks and magnets to make his own

[00:17:51] toys.

[00:17:53] As a boy excited by science and technology he showed an aptitude for mechanical creativity.

[00:17:59] His father, a farmer, sent him to his uncle's place in Surat to study in a good school.

[00:18:05] After class 12 everyone wanted him to be a doctor but he wanted to do engineering and

[00:18:11] took up electronics at the Sarvajani College of Engineering and Technology in Surat.

[00:18:16] In 2008 he moved to Bengaluru to do his masters at the prestigious Indian Institute of Science.

[00:18:24] The Center for Electronics Design and Technology, CEDT at IISC was a perfect fit for Yatin.

[00:18:32] It was a course for those with a passion for hands on work.

[00:18:36] And he found his Methe in the electronic product design and interdisciplinary branch that included

[00:18:43] mechanical and electronics engineering closely linked to robotics.

[00:18:48] Around this time there was a lack of entrepreneurs and industries in the electronic product segment

[00:18:53] and very few tech startups.

[00:18:56] CEDT's focus on preparing students to become creative design leaders in electronic products

[00:19:02] and future entrepreneurs helped unleash Yatin's inner inventor.

[00:19:08] That's where I got my product orientation, not just engineering he says.

[00:19:13] It taught us how to develop a product and how the smallest thing you do can impact customers.

[00:19:19] I became more of a product engineer.

[00:19:22] As a student, Yatin worked on several products including a motion triggered camera trap for

[00:19:28] wildlife studies.

[00:19:30] I did a project on next generation camera traps for counting tigers in the Bandeepur tiger

[00:19:35] reserve he says.

[00:19:37] The camera trap project was originally conceived a few years ago at CEDT.

[00:19:42] Professor H.S. Jamadagni also an avid environmentalist started the project as a hobby, being

[00:19:49] interested in wildlife photography that did not rely on the prohibitively expensive products.

[00:19:55] It soon became more than a hobby and several of his students took part in the development

[00:20:00] and improvement of the camera traps.

[00:20:03] The problem with the earlier version was that it took quite some time to trigger.

[00:20:08] In 2009-10, Yatin worked on a newer version that was very fast.

[00:20:13] It came through because my professor had a lot of experience and we knew what customers

[00:20:18] wanted.

[00:20:19] He points out.

[00:20:21] Gathering experience and moving on.

[00:20:24] Yatin then worked for nearly four years as a design engineer with analog devices and

[00:20:30] MNC.

[00:20:31] He however did not enjoy the MNC work.

[00:20:34] He wanted to do something new that would have a long term effect on the world and finally

[00:20:39] jumped into the startup world.

[00:20:42] In 2014 he joined an early stage startup Leumas design technology to build smart backpacks.

[00:20:49] As a co-founder and chief technology officer, he researched on bringing the next generation

[00:20:55] of variables to life for over two years.

[00:20:59] On the meantime, he acquired a range of patents in the field of variables and led the company

[00:21:06] to successful products and campaigns, raising over $100,000 in Indiegogo, an American crowd

[00:21:14] fund innovation and support enterprise.

[00:21:18] After Leumas ran its course, he took to consulting to fill his time while he decided what to

[00:21:24] do next.

[00:21:26] All the work Yatin had done until then did not use AI technologies.

[00:21:31] In fact, while AI existed as a concept, not much work was happening in the field at that

[00:21:38] time.

[00:21:39] He wasn't to hardware and building world-class products ground up.

[00:21:43] I am a hardware engineer and started thinking if I could really do something to solve

[00:21:47] our food problem, say Ziyatin.

[00:21:51] While in consultancy, he also started making models of his dream project.

[00:21:56] As his confidence grew, he began giving it all his energy and effort from 2018.

[00:22:02] Yatin knew better than many others that although software had been the star of high tech over

[00:22:08] the past decades, the reliance on hardware as a core enabler of innovation was rising in

[00:22:14] all emerging applications.

[00:22:17] As the world generated more and more data each day, the need to develop new hardware

[00:22:22] to support the storm of data was imperative.

[00:22:26] Although they were just a handful of hardware enterprises in India, he knew that hardware

[00:22:31] served as the differentiator in AI.

[00:22:35] Hands on and heads on.

[00:22:38] In the meantime, he discussed his concept with Pranav Ravel, Amit Kupta and Sudeep Kupta,

[00:22:45] his friends from IISC.

[00:22:47] Together they co-founded Euphotic Labs to build Nosh, the AI-based automated cooking robot.

[00:22:55] The result to work together to provide this product to people who needed authentic and

[00:23:00] mouthwatering Indian meals, it took three years of hard work, successes and failures, six

[00:23:07] prototypes, multiple recipe restorations, and countless customer feedback for the team

[00:23:13] to create a solution that could alleviate their culinary walls and offer them with refined,

[00:23:19] pleasurable cuisine.

[00:23:22] For us, AI came as a necessity, explain Siyatin because food preparation is not just a mechanical

[00:23:28] job.

[00:23:29] You also make a lot of decisions when you cook.

[00:23:32] We realize that it's not an electro-mechanical game but one of intelligence as well.

[00:23:38] The team started building their model focusing on the cooking raw materials, the chopping

[00:23:43] sizes and the colour of dishes.

[00:23:46] Nosh was a new category product, as a result they did not have any available reference

[00:23:51] product to learn from.

[00:23:53] They had to develop multiple iterations and experiments to teach the device how to

[00:23:58] give different treatment to different foods.

[00:24:02] Of onions and algorithms, take for instance onion, a staple in every Indian kitchen.

[00:24:09] Should it be golden brown or translucent?

[00:24:12] Which kind of onion should be chosen?

[00:24:16] It was very hard, Siyatin.

[00:24:18] You could get your AI right with onions but what about the things that were to go before

[00:24:23] along with or after onions?

[00:24:25] A recipe may require curry leaves to go with it or rice or spices.

[00:24:31] If you are frying onions for a curry, the ingredients and spices should only be introduced

[00:24:36] after the onions turn golden brown.

[00:24:39] It was a problem that needed a lot of data to solve.

[00:24:43] Hence the team started building many algorithms, the first successful dish was potato fries.

[00:24:49] Boyed by the enthusiasm, they started to scale up for complicated dishes.

[00:24:55] They had to look at individual items, tomatoes, potatoes, poha and figure out how to deal

[00:25:01] with each.

[00:25:03] Their knowledge was based on two things, ingredients and cooking techniques.

[00:25:08] Building the AI was an adverse journey but the team worked on priority and built algorithms

[00:25:14] on the data they had.

[00:25:16] The first algorithm to develop was for the right cooking items.

[00:25:21] Other things like taste, aroma and texture came later.

[00:25:26] Art or chemistry?

[00:25:29] Yatin's mother threw a spanner in the works.

[00:25:32] Cooking is an art she said, you can't do it on a machine.

[00:25:36] Yatin countered with an argument that only scientists and engineers can come up with.

[00:25:42] Cooking is actually chemistry, when you add heat things happen.

[00:25:46] He also argued that cooking is not just a science, it's a natural science because things

[00:25:51] like soil type, the topography, the irrigation all change the ingredients.

[00:25:57] Yatin and his teammates realized that the right combination of time and temperature worked

[00:26:03] well in determining consistency, taste, smell and flavor in most cases.

[00:26:10] Unless the moisture content of an ingredient varied significantly.

[00:26:16] Moisture content determines how long it will take for Nosh to cook, he says.

[00:26:21] That's the key variation in the AI algorithm that we had to deal with.

[00:26:27] So how does Nosh work?

[00:26:29] First up do you need to know basic cooking?

[00:26:32] The answer is no, you really don't need to know anything.

[00:26:38] The first step is to choose the desired item for cooking, you need to load ingredients

[00:26:44] in designated containers according to instructions given on the Nosh app.

[00:26:49] The spices are stored in an inbuilt storage like a spice box at home, vegetables, paneer

[00:26:56] or meat go in the tray, it also has storing units for water and oil.

[00:27:02] Find it all inside the device and command it to cook.

[00:27:06] While the machine does the hard work, you can step out and do your own work.

[00:27:10] Once the food is cooked, you will get a notification on your phone.

[00:27:14] A standard chicken curry usually takes 25-30 minutes.

[00:27:19] Can Nosh do more than just cook your favorite meal?

[00:27:22] Yes.

[00:27:23] It can assist you with meal planning, ordering items from an automated grocer, track calories

[00:27:29] and personalize food to your preferences.

[00:27:33] Going step by step.

[00:27:35] But Yatin's work did not end here, he and his team continued to build prototype after

[00:27:41] prototype to reach perfection and reliability.

[00:27:45] In AI, reliability is a loaded word.

[00:27:49] Simply put it means the application of data analytics including AI machine learning

[00:27:54] to predict when an asset will fail or deteriorate.

[00:27:59] Know that it can be serviced or replaced on time.

[00:28:02] As more data was incorporated into the operating system, Nosh grew more flexible and adaptable.

[00:28:09] The team wanted to go step by step.

[00:28:12] First was user trials.

[00:28:14] They started taking their product to different customer groups to learn from them what worked

[00:28:19] and what did not.

[00:28:21] Funding was another challenge but they were fortunate to receive support from the government of

[00:28:25] India several investors as well as friends and family.

[00:28:29] Recently they raised a preceded round of $1 million by Bitts Park and other angel investors.

[00:28:38] They opened up pre-orders announcing a discount of 10,000 rupees, the price of Nosh is about

[00:28:44] 50,000 rupees for the early bookers.

[00:28:47] Plenty of pre-orders came from within and also outside the country, especially the NRI

[00:28:53] community.

[00:28:54] With about 2,000 pre-orders in hand they began to focus on reliability and delivery.

[00:28:59] Apart from these, they focused on two cities that attract the largest flow of people from

[00:29:04] all across the country, Bengaluru and Delhi.

[00:29:09] A Mother's Sale of Approval

[00:29:11] Yatin's Mother's initial reservation about a robotic chef has been finally put to rest.

[00:29:17] As Nosh churns out complex dishes, poha, upma, biryani, kari paneer, galik prawn, pasta,

[00:29:25] you name it.

[00:29:27] The food also does not taste like something that has not been cooked by human hands and

[00:29:32] it delivers consistent taste every time without supervision.

[00:29:36] Yatin hopes Nosh will make the experience of cooking and eating effortless in keeping

[00:29:42] with one's preference, time and health.

[00:29:45] And he dreams on.

[00:29:47] Perhaps his robotic chef will one day relieve humanity of the hassle of cooking?

[00:29:52] Fingers crossed.

[00:29:54] Part 2 Creating the Sound of Silence

[00:29:58] The story goes back to 2008.

[00:30:01] Eshwarya Karnataki was then a student of class 8 at the Scottish High International

[00:30:07] School in Gurukram.

[00:30:09] The school had a separate section for students with disabilities.

[00:30:13] As a member of the student council, one of her key responsibilities was to interact with

[00:30:18] these children.

[00:30:19] She noticed that while all of them had friends, Atharva, a six year old boy played alone during

[00:30:26] recess.

[00:30:28] She soon found out that he had a hearing impairment and since no one understood sign language,

[00:30:33] he was unable to be part of what was happening around him.

[00:30:37] So Eshwarya started learning sign language and one day asked him, will you be my friend?

[00:30:43] He was ecstatic.

[00:30:46] Now as Eshwarya looks back she recalls how Atharva made her realize that the ability to communicate,

[00:30:54] connect, share thoughts and build relationships is a basic human need.

[00:30:59] Ever since I wanted to do something for those with hearing impairments.

[00:31:04] And that is exactly what she has done.

[00:31:07] Along with Pune based engineer and data scientist Parikshith Millen Soning, who has worked closely

[00:31:12] with people with hearing impairments, she created the AI-powered device 5th Sense.

[00:31:19] 5th Sense helps those with hearing impairments communicate effectively without being dependent

[00:31:24] on a sign language interpreter, she says.

[00:31:27] And it provides equal learning and employment opportunities for them.

[00:31:34] Hearing Assistive Tech

[00:31:36] Hearing loss is the most common sensory deficit today.

[00:31:40] According to the World Health Organization WHO, there are more than 2.5 billion people

[00:31:46] or 1 in 4 across the world who live with hearing impairments.

[00:31:51] Over 430 million of them have disabling hearing loss as per the 2021 WHO Report World

[00:31:59] Report on Hearing.

[00:32:01] AI innovations have the potential to tackle this challenge.

[00:32:05] One sector in particular, AI Assistive Technology or AT has created new buzz.

[00:32:12] AT is a 25 billion dollar global market of specialized products that improve the lives

[00:32:19] of individuals with disability especially in terms of human connection everyday life

[00:32:25] and employment.

[00:32:27] A host of startups are working to develop and deploy new designs with this technology globally

[00:32:33] although 90% of those who need it do not have access.

[00:32:37] Reports the first global report on Assistive Technology, great 2022 of the WHO.

[00:32:45] In India too, the sector has witnessed a significant amount of interest from early stage

[00:32:51] startups and angel investors in the last 5 years.

[00:32:55] An Invisible Disability.

[00:32:58] The advances in AT are transforming many fields but have yet to make an impact on hearing.

[00:33:05] Even with recent advances in AI, there are more devices for those with visual impairments

[00:33:11] than for those with hearing impairments possibly because the former is a self evident disability

[00:33:17] while the latter is invisible.

[00:33:20] The biggest challenge is that millions of deaf people rely on sign language, a mode of

[00:33:25] communication not broadly understood by others.

[00:33:28] What's more, these languages vary from country to country and even in different parts of

[00:33:34] the same country.

[00:33:35] This creates a huge gap in how technology for the hearing impairment can become flexible,

[00:33:40] scalable and sensitive catering to a wide range of people.

[00:33:45] Most of the technology on hearing can augment people's ability to hear sound, hearing aids,

[00:33:52] listening and personal sound amplification products to co-clear implants.

[00:33:58] But these support those who can hear to an extent, rather than those who have total

[00:34:03] hearing impairments.

[00:34:05] Others are designed to alert the hearing impaired about specific events that would usually be

[00:34:11] announced audibly via light, vibration or a combination of the two.

[00:34:16] The third category is communication technologies which deploy keyboards, touch screens or video

[00:34:22] technologies apart from tools to translate speech, text and sign.

[00:34:28] This category is improving rapidly and this is the category that fifth sense belongs to.

[00:34:34] The power of fifth sense.

[00:34:37] Imagine a hearing and speech disabled patient having a private conversation with his or her

[00:34:43] doctor without an interpreter in the room.

[00:34:46] Think of a deaf person giving a confident presentation to a room full of people.

[00:34:52] Consider parents hearing, I love you from their deaf child for the first time.

[00:34:58] All these can become real with fifth sense.

[00:35:01] We wanted to bridge this communication gap and empower hearing and speech impaired people

[00:35:06] to lead independent and dignified lives.

[00:35:09] Mishwaria says she has developed the hardware while Parikshith has built the software.

[00:35:15] Fifth sense has been designed both as smart watches as well as gloves.

[00:35:19] It can convert gestures to speech and text and has an ergonomic design that is comfortable

[00:35:26] to wear, even for long term and sustained use.

[00:35:29] It is also easy to use and set up.

[00:35:33] The device makes for two-way communication with an inbuilt microphone that converts sounds

[00:35:38] to text or image format.

[00:35:40] It also enables the learning of Indian Sign Language through real-time feedback mechanism.

[00:35:46] The ultimate point of fifth sense is seamless communication between sign language users

[00:35:52] and the hearing or speaking community.

[00:35:56] Passion for Startups

[00:35:58] Growing up, the ability to communicate and connect was a vital tool for Eshwaria.

[00:36:03] Her engineer father, currently the managing director of DL Mittels, worked in jobs that

[00:36:09] took him across the country and the world.

[00:36:12] Born in Nasik, she too travelled across the globe and was brought up in cities across

[00:36:18] continents.

[00:36:19] In her first 18 years, she lived in 11 cities, adjusting and making new friends time

[00:36:25] and again.

[00:36:26] Most of her schooling was done in the US, class 10 in Guru Gram, class 12 in Delhi and

[00:36:32] engineering in Manipal, before settling down with her parents in Pune.

[00:36:37] As an electrical engineer from Manipal University which is known for its vibrant startup ecosystem,

[00:36:44] Eshwaria developed a passion for Startups.

[00:36:47] This was my third startup, the 29-year-old says.

[00:36:51] Her first came about in 2016.

[00:36:54] Called MACJ Home Inspection Services, it uses infrared scanning to check the effectiveness

[00:37:01] and quality of a building's construction.

[00:37:04] Her mother and MBA now looks after this startup.

[00:37:08] This second was a facility, ElectroSensei which looked into other transport equipment

[00:37:14] manufacture.

[00:37:16] This startup didn't work but both definitely taught me something, she says.

[00:37:22] Combining her sign language skills and knowledge in electronics, the idea of fifth sense was

[00:37:28] born.

[00:37:29] Her startup takes flight.

[00:37:31] It was in January 2020 when Eshwaria met Parikshith to discuss her idea.

[00:37:37] Parikshith is a mechanical engineer who had graduated from Pune University in 2014.

[00:37:44] He worked various engineering jobs for a few years before realizing that his calling

[00:37:49] was slightly different and started exploring other opportunities.

[00:37:54] That's how I stumbled upon the world of data science, he says.

[00:37:58] He did a six month full-time course on big data and analysis from SP Gen Institute of

[00:38:05] Management and Research in Mumbai.

[00:38:08] For the last five years he worked for multiple corporations in the domain of finance and data

[00:38:14] science, developing extensive experience in predictive analytics, the use of data, statistical

[00:38:22] algorithms and machine learning to identify the likelihood of future outcomes.

[00:38:27] He also taught data science in a part-time capacity at Pune.

[00:38:31] I'm really passionate about data science, says the 31 year old so when Eshwaria told me

[00:38:38] about her idea I was thrilled and what they planned to do had arguably not been tried

[00:38:44] before.

[00:38:45] We actually got the chance to develop the entire technology from scratch, implement

[00:38:50] the AI algorithm and create something new.

[00:38:53] It was a very exciting journey.

[00:38:55] On top of that, the social aspect was immensely motivating for him.

[00:39:00] He two new people with speech and hearing impairments.

[00:39:04] I was instantly hooked on the idea.

[00:39:07] I could see that it would impact many people in a very positive way.

[00:39:12] They registered their company, Glowatrix Private Limited on 1st April 2021.

[00:39:18] As their startup took flight they also got funding commitment from Jito, Jan International

[00:39:24] Trade Organization in May 2022.

[00:39:29] How do I use 5th Sense?

[00:39:31] As a user, you have to wear 5th Sense on both hands because sign language requires both

[00:39:37] hands.

[00:39:39] Once you wear it, you have to connect the device to the app on your mobile phone.

[00:39:44] Then you press the button on the device, perform the sign language gesture and press the

[00:39:49] button again.

[00:39:51] Because the algorithm needs to know when you are signing and when you are making random

[00:39:55] gestures with your hands.

[00:39:57] Then it will speak out in the primary output which is English.

[00:40:02] The device enables two way communication.

[00:40:05] The person with hearing impairment will be able to express what they are saying because

[00:40:10] the device will translate their gesture to speech.

[00:40:13] They will also understand the reply they are receiving.

[00:40:16] The device has a microphone to capture what the other person is saying and convert that

[00:40:21] to text and images on the screen.

[00:40:24] Apart from this, there is a sound detection circuit diagram.

[00:40:29] People with hearing impairments miss out on a lot of audio cues say the dog barking,

[00:40:36] baby crying, points out ashrere.

[00:40:38] New mothers who are deaf suffer from anxiety because of this.

[00:40:43] This device also captures background sound and sends notifications.

[00:40:48] A brand new data soup.

[00:40:50] In the end, everything comes down to data.

[00:40:53] What is data in this context?

[00:40:55] Here is a quick and easy explainer of all the steps ashrere and parikshith have taken

[00:41:00] in the course of developing their startup.

[00:41:04] Parikshith says, let's say you are trying to set a computer vision algorithm.

[00:41:09] So we show the picture of a dog to a mobile camera and it identifies it as a dog.

[00:41:14] In this case, the data needed would be millions of images of dogs for the computer to learn.

[00:41:21] Sounds complicated but in reality it is rather simple.

[00:41:25] There are millions of images out there for doing this.

[00:41:28] A process called computer vision algorithms.

[00:41:31] One can directly download and train the computer.

[00:41:34] For fifth sense, they say the algorithm is the first of its kind and has never been

[00:41:39] attempted before.

[00:41:41] And the data they require is also not exactly out there.

[00:41:45] This has been one of their biggest challenges.

[00:41:48] So how do they collect their own data set?

[00:41:51] The entire technology is based on motion sensing.

[00:41:55] The device has six motion capture sensors for each hand to monitor and map the movement

[00:42:01] of hands and each finger, both linear acceleration as well as rotation when a person does a gesture.

[00:42:09] That allows a real time understanding of how fast the hands and fingers are moving and

[00:42:15] in which direction.

[00:42:16] Each of these movements is matched to a particular word.

[00:42:20] This movement is the data.

[00:42:22] This data is a string of numbers.

[00:42:25] The duo has been giving the device to people with hearing impairments and asking them to make

[00:42:30] each gesture multiple times.

[00:42:33] This enables them to capture variability in that gesture.

[00:42:37] For each word and gesture, therefore there is a series of data that they interpret for

[00:42:42] converting to text.

[00:42:45] The sign of the times.

[00:42:47] Ashwarya and Parikshita are essentially compiling a complete dictionary of such gestures.

[00:42:53] India doesn't have a standard sign language.

[00:42:56] After a long campaign by the Deaf community, however, the government of India has set up

[00:43:01] the Indian Sign Language Research and Training Centre under the Ministry of Social Justice

[00:43:06] and Empowerment.

[00:43:08] Since 2018 it has been actively developing the Indian Sign Language.

[00:43:13] Right now it has agreed to around 20,000 words using which a person can hold a full

[00:43:19] conversation.

[00:43:22] Every year, multiple gestures and words are being recorded and added.

[00:43:27] The current vocabulary size of the Indian Sign Language Dictionary is 20,000.

[00:43:32] That means 20,000 different gestures with N number of terms covering everyday legal,

[00:43:38] academic, medical and technical usage.

[00:43:42] All these terms are already there for them to train in and this has helped the duo to

[00:43:48] solve one of the biggest problems, the standardization of the Indian Sign Language.

[00:43:54] They are also working to incorporate the American Sign Language.

[00:43:58] An algorithm that works.

[00:44:01] Sign language sentence construction and grammar are not exactly similar to English.

[00:44:07] In most cases, the word comes towards the end of the sentence so the machine learning

[00:44:12] algorithm from gesture data will first try to identify most of the words that are being

[00:44:18] said.

[00:44:19] Then these words will be sent into the sentence reconstructing algorithm where they can

[00:44:25] change the grammar and make it suitable for the user.

[00:44:29] Then it will be broadcasted to the speaker.

[00:44:32] This is where AI natural language processing or NLP predominantly used for translation

[00:44:38] from one language to another or from voice to text comes into play.

[00:44:44] In this case however the translation is from the sensor data that is being generated.

[00:44:50] Essentially it is a machine translation approach but from motion sensor data to English text

[00:44:56] happening in real time, a bit like virtual assistant apps such as Alexa, Siri or Google

[00:45:03] Assistant.

[00:45:05] We have been developing for the last two and a half years now, says Parikshith.

[00:45:09] So we have made multiple iterations of the hardware because hardware development is very

[00:45:14] tough.

[00:45:15] We have to try a lot of iterations of the components and research to figure out what

[00:45:20] exactly is working for us.

[00:45:23] Right now we are finalized our hardware, he adds.

[00:45:26] All the functionality that we are proposing in our final devices is already there.

[00:45:31] And we have already manufactured 15 pairs of the devices.

[00:45:36] The beta product.

[00:45:38] Right now the beta product is in two different formats, a smartwatch with strings attached

[00:45:43] to the finger and a pair of gloves with all the wires concealed inside.

[00:45:48] The glove will be open, all the fingertips being exposed so that one can use the fingers

[00:45:53] for everyday activity.

[00:45:55] A screen inside the left palm shows all notifications on background sound such as the door

[00:46:01] bell, cooker whistle, fire alarm and so on.

[00:46:04] The primary language is English but there is unlimited scope for translating it into

[00:46:10] any local language of the user.

[00:46:13] Algorithms for converting English into local languages already exist.

[00:46:17] Work is on to identify words most commonly used in specific workplace settings and develop

[00:46:24] algorithms for those first.

[00:46:27] The initial go to market strategy is to launch the device as a workplace communication tool.

[00:46:33] Fifth sense virtually has no competitors.

[00:46:36] There have been some attempts to solve the problem via vision based technology.

[00:46:41] Google smart classes for instance can translate the American sign language and project it in

[00:46:47] front of a user's eyes like real time subtitling.

[00:46:52] The technology however is not foolproof and must have an ideal setting.

[00:46:57] In our technology we are monitoring the movement of the fingers in three dimensions, in real

[00:47:02] time we don't need any ideal conditions but it should point out.

[00:47:07] The first client's.

[00:47:09] Their first client was Tereson Kitchenenbar, a restaurant on Ferguson College Road in Pune

[00:47:15] which is run by people with speech and hearing impairments.

[00:47:19] They looked for pain points that the staff might be struggling with as well as opportunities

[00:47:24] to solve those problems with the device.

[00:47:27] With menu card showing how to order in sign language it seemed the initial interaction

[00:47:32] went off well.

[00:47:33] Issues however arose when there was any request for customization.

[00:47:38] For instance if you want say pasta without onion, the waiters will not understand, points

[00:47:44] out asheria or if a waiter wants to say they don't have onions,

[00:47:49] they are not able to communicate.

[00:47:51] Hence they had to recruit a hearing manager as interpreters.

[00:47:55] Asheria and Parikshit got in touch with others such restaurants for instance Mirchi

[00:48:00] and Maim and Ishara in Mumbai.

[00:48:03] Everywhere it was evident that waiters and staff members with hearing impairments would

[00:48:08] seldom be promoted owing to their disability.

[00:48:12] We pitched our device and they were all happy.

[00:48:15] Says Asheria, we are working with them now.

[00:48:19] The duo used these experiences to incorporate 300 gestures and restaurant specific words

[00:48:25] using which waiters would be able to hold conversations with their customers.

[00:48:31] They also prepared 26 alphabets to enable the waiters to string sentences together.

[00:48:37] Apart from restaurants they found many other businesses that hire deaf people including

[00:48:42] KFC, Decaflon, PVR cinemas, lemon tree, blinket, miracle courier and others.

[00:48:51] Gaining new insights.

[00:48:54] Developing the device has given Asheria and Parikshit a sharp insight into grievances faced

[00:48:59] by people with hearing impairments.

[00:49:02] For one of their initial surveys they went to the CR Ranganathan College for the Deaf in

[00:49:07] Pune to interact with students.

[00:49:10] This is the only college in Maharashtra where students with hearing impairments can do

[00:49:14] graduation courses.

[00:49:17] We found that students could grasp only 30% of what teachers were saying.

[00:49:21] They need constant eye contact in order to perceive and understand facial expressions.

[00:49:27] In a classroom setting this is particularly difficult as teachers often turn to the board

[00:49:32] and eye contact is lost.

[00:49:34] Says Asheria.

[00:49:36] She also recalls the heartbreaking story of a new mother they interviewed.

[00:49:41] A one day the baby fell and started crying but she could not hear for a long time.

[00:49:47] When she finally took her baby to the hospital the doctor could not understand her sign language.

[00:49:52] It was a nightmare but that sort of trauma is part of their everyday life.

[00:49:58] People with hearing impairments also have a perennial dependence on translators and interpreters.

[00:50:03] While they have to shell out money constantly the presence of a third party is an infringement

[00:50:09] on their privacy.

[00:50:10] Many have bachelor's degrees but end up with low paying manual jobs at companies because

[00:50:16] of their disability.

[00:50:18] They suffer from low self-esteem and need to go for counseling but how will they talk

[00:50:22] to counselors?

[00:50:24] Hopefully our device will help them all.

[00:50:27] The clock is ticking.

[00:50:30] The fifth sense is ready to roll.

[00:50:32] Priced at 25,000 rupees per pair it is slated to reach the restaurants in the next three

[00:50:38] months.

[00:50:39] Business to consumer transactions will follow soon.

[00:50:43] Parikshit is now looking after both hardware and software development while Asheria who

[00:50:48] calls herself a people's person is busy looking after the business side of things.

[00:50:54] Here is hoping the device will create a win-win situation for both investors and the people

[00:50:59] they are trying to serve.

[00:51:01] Part 3 The Cyber Warrior

[00:51:04] Saket Modi has a success mantra.

[00:51:08] It sounds deceptively simple but is singularly hard to follow.

[00:51:13] Seeing things that others miss and doing things not meant to be done that way.

[00:51:19] That is the key that has made Saket India's brightest name in cyber security and ethical

[00:51:24] hacking.

[00:51:25] The co-founder and CEO of Safe Security is today a global name.

[00:51:30] His company leverages AI to protect the digital infrastructure of multiple Fortune 500 companies

[00:51:37] around the world from malicious cyber attacks in real time while measuring, predicting and

[00:51:43] managing safety breach likelihood.

[00:51:46] Based out of Palo Alto, California, Safe Security is considered a pioneer in its space.

[00:51:52] It is also the world's first inside out quantification and underwriting platform to provide

[00:51:58] continuous assessment of cyber risk.

[00:52:01] All of that success had to begin somewhere.

[00:52:04] Born in Calcutta in 1990, Saket got the opportunity to use his way of thinking and possibly

[00:52:10] his first taste of success when he hacked into the IT system of his school to access a chemistry

[00:52:17] exam paper.

[00:52:18] A subject he was not too fond of.

[00:52:21] The teacher had stored it in a password protected word doc and the class 12 student had downloaded

[00:52:26] a brute force cracker free from the internet to gain unauthorized access.

[00:52:32] It, however, turned out to be too much of an ethical dilemma.

[00:52:36] He confessed to the authorities but the sheer joy of being able to do it was too exhilarating

[00:52:42] to forget.

[00:52:43] As morning shows the day one could safely have predicted his future course.

[00:52:49] Those were no enigma to 33 year old Saket.

[00:52:53] His father Sunil Kumar Modi, a long time stock market advisor, had started his journey in

[00:52:59] computer education and multimedia as a business partner to arena animation.

[00:53:04] A training school that offers job oriented courses in animation, multimedia and web designing.

[00:53:12] Saket moved from Calcutta to Jaippur to study computer science at the LNM Institute of Information

[00:53:17] Technology.

[00:53:19] Even as a student, he started giving tutorials, seminars and workshops on ethical hacking

[00:53:25] and cyber forensics which were attended even by doctoral students of computer science.

[00:53:31] The workshops became so popular that Saket started charging a fee to attend premium institutes,

[00:53:37] IITs, IIMs and over 100 colleges in India and outside the country.

[00:53:44] Let's spread the word.

[00:53:46] The passion for raising awareness around cyber security turned into a freelancing opportunity.

[00:53:53] In 2012, Saket set up Lucidius and IT risk assessment and digital security services firm with

[00:54:00] two of his friends, Vidith Bukshi and Rahul Tiagi.

[00:54:05] They were then just 4th year students.

[00:54:08] The trio operated with capital from family and personal savings at the Incubation Centre

[00:54:13] of IIT Bombay, SINE.

[00:54:16] Saket the chief sales officer signed 100 customers in the first two years of operation with zero

[00:54:23] marketing budget.

[00:54:24] He also continued to speak on various IT and entrepreneurship verticals from the Confederation

[00:54:30] of Indian Industry and Data Security Council of India to CEBIT Germany and Cyber Technical

[00:54:37] Israel.

[00:54:39] The year 2016 was particularly notable for hacking incidents in India.

[00:54:44] First, there was massive data breach of millions of debit cards indicating a broader risk

[00:54:49] to the banking sector.

[00:54:51] Then, a group of anonymous hackers which called themselves Legion took over the Twitter

[00:54:57] handles of prominent Indians.

[00:54:59] By this time, Lucidius had major clans including ICICI bank, Standard Chartered Bank, Tata

[00:55:06] Play DTH Services, Indigo and food chain such as Speedza Hut and KFC.

[00:55:13] The companies Clientel also included the ministries of Home, Finance, Corporate Affairs and

[00:55:18] Defence.

[00:55:19] The company also did a security assessment of the Beam Mobile Payment app developed by

[00:55:23] the National Payments Corporation of India.

[00:55:26] Then the Reserve Bank of India asked Lucidius to hack into the IT system of India's central

[00:55:31] bank to identify weak spots and strengthen cybersecurity.

[00:55:36] By 2017, Sarkid had shot to national prominence.

[00:55:41] First for a TV program where he shared a panel with the minister and several top officials

[00:55:46] of the Union IT Ministry and said, your smartphone can be hacked in 20 seconds.

[00:55:53] The nation sat up and took note when he casually asked for a phone from the studio audience

[00:55:58] and accessed its call logs, messages, contacts and GPS history all within 20 seconds.

[00:56:06] The 22nd YouTube clip went viral.

[00:56:10] The limelight shown on Sarkid for two more reasons.

[00:56:13] His company clearly had access to some of the most sensitive data in the country.

[00:56:18] And unlike most other startups in India, Lucidius was already making profits to the tune of

[00:56:24] 1.3 crore rupees.

[00:56:28] AI in Cybersecurity

[00:56:31] Over 2,200 cyber attacks take place each day.

[00:56:36] A mid-sized company gets alerts for over 200,000 cyber events every day.

[00:56:42] Over 560,000 new pieces of malware are discovered a day.

[00:56:48] There has been an 87% increase in malware infections over the last 10 years.

[00:56:54] Malware attacks cause an average loss of 50 days in time for businesses.

[00:57:00] The average cost of a data breach globally has hit $4.35 million.

[00:57:07] As cyber attacks grow in volume and complexity, AI in Cybersecurity is helping improve how security

[00:57:15] experts analyze, study and understand cyber crime.

[00:57:20] Sarkid's pitch to the chief information security officers of companies focuses on this.

[00:57:26] Nothing is more important than quantifying your risk.

[00:57:31] AI with its incredible capacity to harness data can easily cut through the noise of millions

[00:57:38] of daily cyber incidents to curate and quantify threat intelligence, recognize data patterns,

[00:57:45] identify new malware, generate alerts and protect sensitive data for organizations.

[00:57:52] The threats that business networks face are not constant.

[00:57:56] They change as attackers frequently change their tactics.

[00:58:00] This makes it difficult to detect every variant of malware and prioritize security

[00:58:06] tasks for cyber personnel in a company.

[00:58:09] Machine learning is ideal for drawing on data from any type of malware

[00:58:14] that has previously been detected.

[00:58:16] As a result, when a new type of malware surfaces, the system may compare it to the database,

[00:58:23] analyze the code and prevent the attack.

[00:58:26] This method works even when a dangerous code is concealed within enormous volumes

[00:58:31] of innocent code. A network monitoring tool powered by AI can track what users do regularly,

[00:58:39] spot anomalies and react accordingly by evaluating this data, a major advantage in a continuously

[00:58:46] changing world. Silicon Valley to Safe

[00:58:50] Lucidius scans IT systems of its clients for possible hacking or malware attacks

[00:58:57] and suggests remedies and services if the systems are found to be vulnerable.

[00:59:04] Early on, Lucidius' vision was to work on digital risk.

[00:59:08] As the company won funding from venture capitalists as well as grew its revenue,

[00:59:14] the Safe Product line came into existence.

[00:59:18] Safe, which stands for Security Assessment Framework for Enterprise, could assess risk across

[00:59:25] people, processes, technology, products and customers helping with prioritized decision-making.

[00:59:32] A client could now view the real-time status of its IT framework on a smartphone and monitor potential threats.

[00:59:40] In 2018, Sakeeth had a chance encounter with John Chambers, Silicon Valley Luminary

[00:59:46] and former CEO of Cisco, one of the largest technology companies in the world,

[00:59:52] as well as CEO of JC2 Ventures, which helps disruptive startups around the world build and scale.

[01:00:00] Soon, JC2 led an investment round that pumped $4 million into Lucidius. Sakeeth moved to the

[01:00:08] US and his co-founders remained in India. Chambers became his mentor, setting the culture of his

[01:00:15] company, Product Vision, go-to-market strategy and more. In 2019, Lucidius raised $7 million

[01:00:24] in venture funding. The fledgling company became India's highest valued cyber security startup,

[01:00:30] with triple-digit growth and more than 500 employees in the US, India and Singapore.

[01:00:37] The Safe Platform has made the company a game-changer in the cyber security space.

[01:00:42] A completely new category of product that makes cyber security threats,

[01:00:47] free from technical jargon, helps create a cyber-risk scorecard for the IT system

[01:00:53] and allows the CEO to make informed decisions.

[01:00:58] It's SafeMe mobile app empowers consumers to easily monitor and understand

[01:01:04] their cyber risk across all the devices and digital platforms they frequent,

[01:01:09] such as Google, Twitter or Facebook in real time. It also includes short videos and quizzes

[01:01:16] on safe practices to follow while using social media platforms against email and SMS scams

[01:01:23] and other apps that compromise one's privacy. Taking on the hackers,

[01:01:29] the last few years have seen multiple Fortune 50 brands procuring Safe

[01:01:34] and a slew of awards coming in but fighting cyber crime has no room for complacency.

[01:01:41] Hackers are no longer 15-year-old nerds sitting conspiratorially in a garage with a laptop

[01:01:48] trying to identify weaknesses in a network, crack an algorithm and gain unauthorized access.

[01:01:55] Hacking is now a fast professionalising industry often sponsored by states.

[01:02:01] Criminal groups and terrorists are also some actors out there. They are all well-funded,

[01:02:07] equipped and trained. So what does Lucidius do? With rebranding in 2021 from Lucidius to safe

[01:02:14] security, Sarket has distanced himself from the hacking industry. The new name indicates

[01:02:21] branding from services to products. Sarket has also moved into cyber security training

[01:02:27] as a way to build a pipeline of cyber security professionals across industries who are in short supply.

[01:02:34] The startup has recently tied up with several universities in India to start post-graduate

[01:02:40] diplomas in cyber security and law. We want to create an army of cyber warriors,

[01:02:47] the engineers of tomorrow say Sarket and it's no surprise considering how staying one step ahead

[01:02:53] of the crowd has always been his USB. Igniting ideas for impact is an audiobook published by

[01:03:01] the Better India in association with Accenture India. Find out more about Accenture India's initiatives

[01:03:08] at www.accenture.com slash IN-EN. Author Damyantidatta, edited by Divya Setu and Tanya Singh.