Generative Artificial Intelligence (GenAI) has fast become a top strategic priority for technology services companies. According to a recent EY study, 55 per cent of Indian tech-services companies’ CXOs report that GenAI will have a high impact on their businesses; 15 per cent state that the implications are existential. While executives believe that the technology provides new opportunities for revenue growth, cost reduction and productivity improvement, only a third of the respondents rate their organisation’s readiness to benefit from GenAI as “high.”
Adoption challenges include unclear use cases, converting pilot projects into production-grade engagements, and unwillingness of many customers to sign up given concerns around accuracy, misinformation, bias, security and privacy. This has led some leading experts to suggest that GenAI’s potential may have been overhyped and may soon push stakeholders into a trough of disillusionment.
This assertion, however, is unwarranted. Like previous disruptive technologies, GenAI will realise its true potential only through experimentation and iteration. However, even as the technology evolves and matures, tech-services companies can focus on key opportunity areas that will serve as building blocks to harness the long-term value of GenAI.
Data governance – Good governance enhances data usability, reliability, security and regulatory compliance. It also focuses on privacy protection, given that personally identifiable information and intellectual property can find their way into large language models (LLMs) through training data or through text generated from user inputs or prompts. Well-governed training data enables repeatability, explainability and greater value-extraction from GenAI use cases as well. Unfortunately, less than 20 per cent of executives trust their enterprise data, and less than 10 per cent of companies have the data and cloud maturity required to deploy AI at scale. These present key opportunities for technology services companies.
Specialised models – Organisations seeking to leverage GenAI for addressing sector-specific challenges are evaluating the costs and benefits of open-source models trained and fine-tuned on their internal datasets. They often need help in building private LLMs which offer contextual relevance, or in building small language models that cost less and produce comparable quality content for narrowly defined bespoke solutions. For example, a tech services provider is assisting a global bank in leveraging GenAI for fraud prevention by early detection of deepfakes in the know-your-customer process. In addition, some tech companies are building proprietary models trained on domain-specific data residing in their knowledge repositories. They are also engaging in discussions to create Large Action Models that transform LLMs into actionable tools capable of completing tasks, including handling variations, without human intervention. These and other IP-led solutions, accelerators and assets can be monetised by the tech companies.
Delivery productivity – GenAI can significantly accelerate the software development life cycle, especially across the development, testing and deployment phases. Early pilots suggest that the value generated can be significant: 20-50 per cent for application development and support, 20-30 per cent for infrastructure managed services and 40-50 per cent for application modernisation and cloud migration. In addition to helping organisations do more with less, GenAI can amplify human potential by freeing up time for creativity and innovation.
Customer experience – Despite the proliferation of AI-powered chatbots that work alongside human agents, many contact centres struggle to improve customer satisfaction scores. High attrition rates impede sustainable performance improvements. GenAI can offer a better self-service experience through the LLMs’ advanced multi-modal classification and summarisation capabilities. This can be a game-changer for personalised customer interactions, knowledge management and content creation. Consider the case of a GenAI-based business process management (BPM) company in the US. It offers customer support which can have seamless phone interactions, lasting over thirty minutes, and sound like a real human. With the ability to remember contexts, it can perform tasks across hundreds of applications automatically, without any training or supervision. It is not surprising that BPM players are reinventing themselves on a war footing.
Analysts expect GenAI-related enterprise spends to grow exponentially over the next decade. Realising this opportunity would require technology services players to rapidly develop relevant market offerings, revisit delivery models and talent priorities, and ensure that risk management, ethics and trust are at the heart of all GenAI initiatives.
Nitin Bhatt is the Technology Sector Leader at EY.
Disclaimer: These are personal views of the writer. They do not necessarily reflect the opinion of www.business-standard.com or the Business Standard newspaper
/cdn.vox-cdn.com/uploads/chorus_asset/file/25290333/STK255_Google_Gemini_C.jpg)
