With over 15,000 employees in over 27 countries, Nagarro is a leading global software development industry with its technology and software consulting services. Last month, the company announced a huge global hiring spree and new centres in India. Nagarro is at the forefront of the AI disruption with its AI accelerators, data-driven business insights, big data, chatbots and more. Analytics India Magazine got in touch with Anurag Sahay, Managing Director and Global Head – AI and Data Science at Nagarro, to learn more about the company’s AI growth, the direction of the industry and becoming a Nagarrian.

AIM: What is hyper-automation, and do you think it is the next big thing for enterprises? 

Hyperautomation, in layman’s terms, is a business-driven approach widely utilised by businesses to identify and automate complex tasks and processes that usually depend on human inputs and knowledge. It involves the systematic use of cognitive technologies such as Artificial Intelligence (AI), Machine Learning (ML), Machine Vision (MV), Natural Language Processing (NLP), etc., which enables streamlining business and IT processes, thereby bringing scale and speed to workflows.

According to Gartner, by 2024, organisations will lower operational costs by 30% by combining hyper-automation technologies with redesigned operational processes. It can vastly reduce human intervention errors in relevant areas, improve cost efficiencies, and enhance consistency and process quality with faster turnaround time. With a range of tools like Robotic Process Automation (RPA), AI, and ML working in harmony to create a cohesive automation suite – hyper-automation is a means for real digital transformation. With its holistic administration of operations and iterative approach, it will give companies the competitive advantage they need to sustain and thrive well beyond 2022.

AIM: What are all the innovative ways Nagarro is using to provide an extra edge to its clients in the competitive market? 

Nagarro brings state-of-the-art AI technologies to its clients to implement hyper-automation, which includes expertise in supervised learning, reinforcement learning, and generative technologies. 

We bring a human-centric design philosophy to implement hyper-automation initiatives. This is important because successful initiatives are critically dependent on an augmented approach to automation rather than a blanket approach. We also operationalise and build complete Continuous Integration (CI)/Continuous Delivery (CD) and MLOps pipelines, as well as bring best practices around data, API, and ecosystem integration.

AIM: The amount of data is increasing at an enormous speed, but the businesses fall back when it comes to taking full advantage. How does the team at Nagarro ensure optimum use of this data? 

Sensors, wearables, and IoT are important data sources that can be leveraged to gather more insights and change the way we conduct things. Self-driving cars, industrial automation, medical diagnostics, supply chain, etc., are a few areas where this impact will be widely felt. At Nagarro, we work towards harnessing the power of unstructured data, as well as leveraging cloud technology to view segmented data in a much more flexible and transparent way.

While the amount of data generated is steadily increasing, the quality of data remains a challenge. Nagarro pays special attention to data quality frameworks and uses them to make data-specific and relevant implementations. Apart from the on-time availability and quality of data, it must also be put in context before it can be harnessed efficiently. Again, this is a domain-driven exercise that we understand and actively work on with our clients. 

AIM: Companies are leveraging AI rapidly. Can you suggest some of the most critical areas where AI will play a dominant role in the near future? 

Analytics, automation, and hyper-automation are key areas where AI is creating an enormous impact and will continue to do so in the near future. The types of data which can be used to design analytical solutions is very broad now, and AI is a catalyst for this activity. AI is also playing a key role in predictive modelling and optimisation, which is used critically in all kinds of decision support systems in an enterprise. Human-machine interaction is another impact area for AI and is changing the interfaces and modalities of how we interact with machines.

AIM: As the head of AI and data sciences, can you share some of the difficulties you face while leading a team of data scientists? 

Hiring has been a key challenge area. The pandemic has largely affected the talent supply and skill gap across India, specifically in key technical roles. A recent report by Nasscom-Zinnov highlighted that India’s current tech workforce comprises a total of 47 lakh employees (2021), while the country needs 52 lakh tech professionals. From a percentage of supply standpoint, this translates to a tech talent gap of 21.1% – the lowest among all the leading economies globally. Moreover, it is estimated that our country will be dealing with a shortage of 14-19 lakh tech professionals by 2026. AI, big data analytics, IoT, and cloud computing are the most in-demand skills, but also faced a huge shortage of workforce in 2021, with a shortage of 1.5 lakh, 80,000 and 1.7 lakh domain professionals, respectively. 

While finding the right talent pool is critical, reskilling and upskilling the existing workforce amid the evolving and dynamic environment is also crucial. It is a major challenge to overcome. Assessing the employees’ existing skill sets and competencies, mapping them with the current requirements, and identifying what gaps to bridge can help overcome this difficulty. Data silos is another challenge – different departments tend to store data at different locations, which makes it difficult to track and stitch it together for further utilising it to develop effective solutions.

AIM: Coming to the hiring process, what all technical skills do you look for in a candidate joining your data science team?

We look for fundamentals that ensure the optimum delivery of any application. In terms of basic prerequisites, we look at mathematical proficiency, especially in Linear Algebra, Probability, and Statistics, and the candidate’s ability to use ideas of Data Distributions and Hypothesis Testing. We also focus on the candidate’s competence in Data Visualisation using the Python/R data science stack and their ability to design and implement workflows of Linear and Logistic Regression and Ensemble Models (Random Forest, Boosting) using R/Python. More importantly, we gauge if the person will be a good cultural fit for the team, is a team player, and has good communication skills.