One look around and we find applications of AI everywhere, be it finding insights from mountains of data or automating production and deliver services. AI is being embraced by an increasing number of businesses, individuals and even governments to boost productivity and raise efficiency. At a country level, while most countries have just started using AI, the technologically advanced countries have considerably leaped ahead. In this context, it will be interesting to find out India’s vision regarding AI and the growth consequences attached to it. 

1. AI to play a significant role in realizing India’s Vision for 2025

The bold vision to make India a $5 trillion economy by 2025 can be achieved through adoption of AI 

Data and AI could add $450-500bn to India’s GDP by 2025. Close to 45% of this value is likely to be delivered by 3 sectors: consumer goods and retail, agriculture and banking and finance. In the agricultural sector, AI can help double farmers’ income through improved production planning and yield. While in the BFSI sector, AI can assist in financial risk modelling and credit underwriting, in the consumer goods and retail sector, it can provide personalized campaigns and targeted marketing. 

However, to recognize the full potential of AI in a sector, an important factor to be considered is the structural composition and the technological maturity of the sector. In general, digitized sectors and firms are more likely to adopt AI than their peers who lack the necessary infrastructure for adoption. Some of the deterrents that stand on the road to adoption are cost of technology, lack of skilled talent and legacy systems which affects the efficiency and quality of outcomes. The journey towards AI maturity can be best described from the diagram below.

Explorer: The new-entrant who is figuring out ways to use AI

Enthusiast: Go beyond proof-of-concept to drive business-led and IT-enabled execution model

Expert: Collaborate with industry disruptors to use AI across the organization

Evangelist: Concentrates on democratization of AI and ensures that outcomes are aligned with business metrics

The ability of AI applications to predict across a range of tasks has tremendous implications. A growing number of start-ups with their unique ideas have translated this into reality

Economic Implications

• Reduction in turnaround time and costs
• Increased efficiency in supporting functions Ex: Nebulaa (with the help of AI and data from various agricultural markets in India for quick, accurate and cheap quality testing) Ex: Aspiring Minds (blends AI with psychometric tests to help employers screen potential employees more efficiently)

Social Implications

• Early prediction of certain diseases such as TB, lung and breast cancer
• Regulating critical natural resources such as groundwater and renewable energy. Ex: Artelus (uses AI-backed screening tool that uses deep learning to check for Diabetic Retinopathy, early onset of TB, lung and breast cancer) Ex: Vassar Labs (predicts groundwater rich areas and the rate of groundwater depletion using data from sensors and satellite as inputs to local governments)

Labour Market Implications

• AI-based solutions creating narratives similar to human

2. Deep-dive into sectors: Retail and Healthcare

RETAIL

The Indian retail industry is undergoing momentous transformation on the back of changing consumer behaviour and adoption of new technologies by businesses. Given the fact that the Indian retail sector is largely unorganized, AI can play an important role in plugging in the gaps. 

Key points:

• The retail sector in India is set to double by 2024 to touch $1.4 trillion from the current $790 billion
• AI-led disruptions in business functioning will aid in the expected ~3X growth in organized retail and e-commerce. The unorganized sector, too, will grow by 40%
• The COVID-19 pandemic is pushing the need for digitization and automation, thereby increasing reliance on AI
• Top priority areas for AI implementation include: customer experience (customer feedback, customer service chatbot), in-store and online operations (product personalization, shopping assistant) and logistics and distribution management (automated product sorting, delivery agent assistant)
• 90% of retailers who have implemented AI solutions are not fully satisfied. Trust is seen as a major hindrance in AI adoption. Also, existing governance and risk management policies for AI are considered inadequate
• Collaborating and partnering with AI providers will help solve implementation bottlenecks and built right capabilities for AI implementation
• More than 80% of retailers believe that reskilling existing talent with necessary AI-related skills will be effective

HEALTHCARE

The Indian healthcare industry is expected to witness a strong growth backed by robust government and corporate investments. India, with its huge resources of unstructured medical data and population diversity, combined with the vast pool of human talent and mushrooming health-start-ups is perfectly positioned to embrace large-scale AI implementation. 

Key points:

• Despite several challenges which are further fueled by the pandemic, the healthcare sector is still expected to reach $372 billion by 2022
• Government-led expenditure towards the sector will increase from 1.6% of GDP in 2020 to 2.5% by 2025
• AI can be effectively used to digitally transform the sector thereby resulting in affordable treatment and improvement of quality and accessibility of services
• Top priority areas for AI implementation include: patient care and experience (personalized treatment, smart wearable sensors, patient assistance chatbots), operations (appointment assistant, hospital patient flow management), R&D (drug development, disease simulation, drug repurpose)
•  60% of healthcare organizations are not satisfied with AI implementation. More than 70% believe hiring new talent to build AI capabilities is effective
• Dedicated AI strategy and budget is a key imperative to scale AI initiatives
• Need for robust platforms and solutions with right capabilities for successful AI implementation
•  Long-term partnership, shared risk models works best with AI providers in the healthcare sector. Special focus on health-care start-ups

3. Building blocks to promote AI and data utilization across sectors

Countries promoting data utilization and AI are doing so based on 5 building blocks

Strategy: 

• Identify priority use cases based on feasibility and impact potential
• Prepare integrated action plan
• Create a long-term implementation plan and funding mechanism

Data:

• Identify data sets required to solve priority use case
• Set standards for data collection, classification and security
• Design programs to generate data and derived services on a large scale

Technology Stack:

• Create a compatible platform to securely host data, AI services, models, open-source libraries
• Create enabling infrastructure to support the ecosystem (ex: 4G, 5G connectivity, sensors)
• Formulate policies to ensure security, reliability and interoperability of the stack

Talent:

•  Define AI specific roles such as data scientists, data engineers and spell the required trainings and certifications
• Estimate demand-supply gaps in AI workforce
• Develop strategy to build the gap and upskill existing talent

Execution:

• Design a national program for AI with clear structures, roles and processes for various stakeholders (government, industry, academia)
• Drive innovation and change management
• Set up an independent body to enforce AI policies

4. Immediate actions required to improve data utilization and AI in India

6 coordinated actions over the next one and a half years required to catalyze the utilization and adoption of AI

• Launch the National Program for AI and create a central and apex body consisting of representatives from various stakeholders for its execution
• Finalize India’s Data and AI Action Plan and develop a self-sustaining financing model for the future
• Based on selected initiatives, identify datasets of national importance (ex: healthcare data, farm data, weather data, land records, education, power and grid)
• Initiate work on a few socio-economic programs which requires scaling up
• Create schemes to engage the AI ecosystem which consists of industry, start-ups, academia and the civil society
• Provide access to datasets of national importance through reliable technology platforms which will act as a marketplace for both private and public sector

5. A look at the Technology Stack and AI platform

An AI platform acts as an accelerator to support quick to build solutions and services for enterprise adoption. The platform should not only ingest and manage data but also generate models. AI platforms are the future of technology services delivery where Indian technology companies have already exhibited strong AI capabilities and successfully developed full-stack of AI platforms. Even though the current state of AI platforms is considered as maturing, the Indian companies are catching up with their global counterparts in terms of complexity and enterprise-wide deployment. 

Key Points:

• The AI platforms have served more than 500 clients across the globe
• Have served more than 15 sectors
• Have a collective platform experience of more than 20 years

Sector Adoption (%) of AI Platforms

Maturity of AI Platforms

Advantages of AI platform

• End-to-end AI Model lifecycle and workload management
• Address various challenges relating to system failure and issue resolution
• Build quick solutions for enterprise use-case
• Extract, analyze and use data to make meaningful insights
• Offer solutions that are easy to consume and govern across enterprises

Interesting Use-cases of AI Platforms