Author Profile

Prabhjinder Bedi

Chief Growth Officer, Business Process Services, Tech Mahindra

Prabhjinder Bedi

Bedi has over two decades of experience involving launching start-up ecosystems, scaling up businesses, and successfully taking products and services to market across industry verticals, spanning telecom and media, hi-tech/new economy, financial services, retail and consumer goods, manufacturing, and life sciences. Having spent over 16 years at Tech Mahindra, Bedi is currently responsible for taking our existing and new-age service offerings to global markets and adding meaning to our shareholders, partners, and customers. He holds a bachelor’s degree in engineering from IIT- BHU and a Master of Business Administration (MBA) degree from IIM Calcutta.

Insights

Views

Most organizations implementing AI into their operational workflows eventually realize that the pilot-to-production transition is a formidable challenge. This holds true even with strong executive support, solid research, expert data science teams, approved tools, and promising use cases. A 2025 MIT NANDA study found that about 95% of enterprises are getting zero return on their Gen AI investments, largely due to the sheer complexity of translating a promising test phase into a reliable business outcome that involves real-world variables.1

Views

As enterprise adoption of AI accelerates, governance is now just as important as innovation. Enterprises are chasing every use case they can find, while regulators struggle to build a governance framework to manage the associated risks. The European Union Artificial Intelligence Act (EU AI Act) is the first attempt at closing that gap. This comprehensive legal framework sets the tone for how enterprises can build and use AI systems. Simply put, the Act protects the public from AI risks, keeps the technology human-centric, and demands real transparency into how AI systems collect, process, and use users' data.

Views

Enterprises have been experimenting with generative AI for a few years now. The conversation started with one simple promise: AI can help people work faster. That led to the rise of GenAI “copilots”, the tools that could summarize long documents, draft emails, support customer service agents, create reports, or help developers write code. The promise was fulfilled; these tools have improved personal productivity and helped employees reduce the time spent on simple and repetitive tasks. But as organizations started using them in real business environments, one limitation became clear: delegating individual tasks to bots does not necessarily complete that operation.