Cloud and AI specialist, Shreyak Gupta, has helped startups and fortune 500 companies alike in futurifying their technological needs. He is an MBA by education and an engineer at heart. Shreyak holds multiple professional and specialty certifications in the field of Cloud, Data and Machine Learning. He aims to use his skills in understanding customer problems and solving for better. While off work, Shreyak likes to automate his daily chores, paint and spend time with his family.
On a Friday afternoon, at a company’s cafeteria, there were sounds of clapping and whistling and a happy birthday jingle echoing from every corner. Ankur was turning 30 over the weekend and his colleagues decided to have a small celebration. During the festivities, Shreya approached Ankur and asked him if he had already bought a life insurance policy. Ankur was baffled by such a thought, especially while he was celebrating life. Shreya further explained that if he bought life insurance before 30, he’d enjoy the perks at a lower premium. Ankur was now on the clock. There were merely hours before he’d turn 30 and he knew these processes took time. He tried connecting with his agent, Prakash but got no response. Taking matters into his own hands, he did some policy comparisons and decided to go ahead with a term plan offered by one of India’s leading insurance providers. He predicted it to be a long process with multiple documents to be filled in, and documents to be uploaded before they get processed and approved by the company. With the weekend a couple of hours away, Ankur lost hope of the process being completed before Monday. He filled in all the required information and waited. Within a few minutes, his documents were processed. Surprised by the swift service, Ankur imagined if there was a guardian angel just waiting to process his request.
On Monday, he called up his agent and told him how the process seemed like magic. Prakash explained how they’ve incorporated an AI called Recognic, which leverages the power and scalability of Google Cloud. How this AI intelligently processes thousands of documents in mere seconds, extracts and validates the required information and lets the user know if any document is incorrect or missing. While satisfied with faster processing and decreased wait times, Ankur thought about how the new age technology is changing our lives for the better.
The talks of Artificial Intelligence (AI) might seem like a dream of a dystopian future, but it has become an integral part of our present. From Google Home controlling our favorite music to Amazon recommending the product of our choice, artificial intelligence is silently present all around us. In the last decade, AI has seen multifold growth and there has been a 450% upswing in Applied AI jobs.
Tech innovation transforming our daily lives
In India, a lot of AI startups have sprung up that focus on solving complex business problems. Several delivery services and cab aggregators leverage intelligent route optimization to figure out the fastest way out while showing the driver’s ETA to reach your location.
The terms “frequently bought together” or “Other people also liked” are not foreign to anyone who has ever been on an e-commerce platform. The retail industry heavily relies on AI-based recommendation engines. The objective is simple – increasing the overall cart value. What’s interesting to note is that something that seems so trivial has a lot of driving potential and has seen an investment upward of $2.5 billion globally.
Route optimization is another ML algorithm used in many delivery and cab aggregating apps. From Ola and Uber showing the shortest route; to Swiggy and Zomato assigning nearby orders to a delivery executive, we see multiple examples of AI enabling these tech companies.
Disruptive trends that are set to rule the industry
Innovation has been the buzz word in every Industry that exists. Technology has played a pivotal role in ensuring businesses evolve and move with the pace of changing customer expectations. Recently, the industry has been seeing the below trends emerge as strong support suits.
- Digital twins. They are virtual copies of a physical thing. There has been a recent uptrend in applications of digital twins in the metaverse. They can help simulate real world shopping. While sounding futuristic, the earliest application of digital twins was 50 years ago, used in NASA’s Apollo 13 mission. These simulators were one of the most complex technologies of the mission. Digital twins have come a long way from there and now see a lot of industrial applications.
- With the advent of OCR (Optical character recognition), Doc AI[6] and other Vision based AI solutions, physical documents are now digitized with ease. Some companies that have successfully implemented Doc AI have cut down the middle and back-office work, leading to a 30% increase in the number of applications that can be processed in the same time span without the need for additional resources.
- A big revolution in the AI space has been in the world of conversational AI. A chatbot can be seen greeting us everytime we visit any website, present 24×7 to solve any query. The IVRs are also developed to point the user in the right direction by simply answering a couple of questions. As these technologies develop further, the chances of talking to a real human seem bleak.
- Structured chats and call transcripts can also enable an AI to classify future interactions into a predetermined category. This helps in understanding the customer better.
- Having a highly available AI run, the show allows multiple users to interact with the system without any wait times. This also helps in keeping human intervention to a minimum thereby bringing down the operational costs and elevating the customer experience. Any call center still following traditional ways without AI, is set to die out sooner than later.
- With the advent of MLOps, experimenting with ML models and translating these experiments to production became easier thereby reducing technical debt across machine learning models. Today, many open source frameworks such as Kubeflow and TFX help in making portable and scalable ML workflows and reduce the bottlenecks that traditional ML models face.
Challenges the industry is yet to address
While all these use cases fit the bill for artificial intelligence, real-life implementations have always been a challenge. According to a survey by Algorithmia, around 55% of the companies that actively work on AI/ML could not deploy a single model[2]. The staggering numbers show that there is clearly a gap in the market where data scientists are not able to engineer pipelines, operationalize models and build scalable solutions that could handle incoming traffic. Companies have started identifying this problem and have started hiring more and more Data engineers. In 2019, the jobs for Data engineers grew by 50% (data scientists grew by 32%).
Recently, there have also been many controversies around the biased and unethical use of AI. Personal data being siphoned for targeted advertisements by Big Techs, examples of racial and gender-based bias in the ML Models are just the tip of the iceberg. Though there has been a lot of movement in the governance of AI and Ethical and Responsible AI has come into practice in the western world, this is yet at a very early stage in India.
Corporates have started figuring out solutions to these problems. Firms have started setting up in-house teams of data engineers and data scientists along with consulting partners who work in tandem to leverage technologies such as cloud to build scalable solutions and implement MLOps to frequently retrain their models.
While we have come a long way in the adoption and implementation of technology there is still so much to be done. The next 5 years are going to be critical as we race towards digitizing businesses and our economy. Positive steps taken today will pave the way for our future.