Decoding Gen AI, Cloud, and VDI: A Candid Conversation with Rajiv Ranjan Kumar of Wipro

Hans India

4 days ago

In an exclusive sit-down with The Hans India, Rajeev Ranjan Kumar, a GenAI specialist and leader, demystified some of the most transformative technologies of our time: Generative AI (Gen AI), Cloud Computing, and Virtual Desktop Infrastructure (VDI). From the promise of AI to its potential pitfalls, and his personal journey into the tech world, Rajeev offers clear-eyed insights with relatable examples.
Here's a glimpse into an insightful chat with him.
Why are companies increasingly embracing Gen AI-powered solutions?
Rajeev explained that the industry is witnessing a major shift towards the convergence of data forms; text, video, image, and audio. Gen AI leverages all of these together, making it vastly more efficient and intelligent. He pointed out how tools like ChatGPT offer a stark contrast to traditional search engines. Instead of returning a list of options, Gen AI delivers precise, context-aware answers, improving productivity and saving time. According to him, this convergence is what's driving the mass adoption across industries.
How are Cloud and VDI connected to Gen AI in today's tech ecosystem?
To break it down, Rajeev categorized AI into three key layers: infrastructure, cloud, and access. First, the infrastructure layer is crucial because AI models need high computing power. But that kind of investment isn't feasible for smaller organizations; which is where cloud platforms come in. With cloud's 'pay-as-you-go' model, anyone- even students can access powerful AI tools without buying expensive hardware. Then, comes VDI (Virtual Desktop Infrastructure), which allows people to securely access their workspace from any device, anywhere. Together, these elements form a robust ecosystem that makes AI scalable and democratized.
What are the biggest challenges in adopting Gen AI?
Rajeev outlined three major concerns: data privacy, ethical use, and hallucination. In sectors like healthcare and automotive, data sensitivity is extremely high. Any leak can lead to serious consequences, including loss of trust or business strategy exposure. Ethical use is another concern, especially with the rise of deepfakes and voice cloning. Rajeev stressed the importance of governance frameworks and audits to ensure responsible use. Lastly, he pointed to the problem of hallucinations, when AI outputs something that seems accurate but is factually incorrect. In high-risk industries, even one such error can be catastrophic.
Are AI-powered vehicles, like those from Tesla, really safe?
On the topic of autonomous driving, Rajeev admitted that the adoption rate remains low, primarily due to data reliability concerns. These vehicles rely entirely on AI models for decision-making, and if even a single command is wrong, the outcome can be dangerous. Hallucinations in AI; where answers look accurate but aren't, are especially risky here. This is why full automation is still under cautious implementation.
Will AI eventually replace human jobs?
Rajeev acknowledged that AI will partially replace roles, particularly in areas like technical documentation and basic coding. Generative AI can produce text and code with impressive accuracy, reducing the number of people required for such tasks. However, he emphasized the continued importance of the human-in-the-loop approach. AI still lacks instinct and random human judgement, and will take years to truly mature. 'You can think of it as the 'AI-fication of humans' already happening, but the 'humanification of AI' is still far off', this was established during the conversation.
Does frequent use of AI tools hamper human creativity?
'No, it actually enhances creativity,' Rajeev said firmly. He shared a story from a gastroenterology summit, where a doctor failed to diagnose cancer in a patient early on. Years later, when the patient's previous records were uploaded into an AI system, it accurately predicted the cancer risk that was overlooked. The experience reinforced Rajeev's belief that AI complements human effort and helps professionals work smarter.
To quote Rajeev,"The patient had a second stage of cancer. The doctor felt guilty. The reason that the person had come to him three times and he was not able to diagnose that he is developing that thing. Then he was feeling guilty and then he told this person to give him all the current applications he has which are based on artificial intelligence. He uploaded the entire data. The first year data, first year data was showing that there is a probability of 50% cancer happens after two to three years. Second year data was telling that 70% probability is there that he will be diagnosed with cancer in the next one year. So, he was surprised that okay, he has so much experience and this fellow came just now as a technology and it is replacing me. But he was thinking I could have used this technology two years back and I could have saved his life much earlier."
In creative fields like media, he said AI can fast-track execution, allowing professionals more time for vision and innovation.
What about the environmental cost of training large AI models?
Rajeev acknowledged the concern about AI's water and energy consumption, but said the industry is responding. Companies like NVIDIA are creating more energy-efficient hardware and Small Language Models (SLMs) are emerging as lightweight alternatives to Large Language Models (LLMs), consuming less power with similar performance for specific use cases. He added that countries like those in the EU are already implementing Responsible AI frameworks, which include environmental considerations.
How do SLMs compare to LLMs, and where should each be used?
SLMs, according to Rajiv, are ideal for task-specific applications such as call centers or IT helpdesks, where the questions are predictable and datasets are limited. LLMs, on the other hand, are better suited for complex, multimodal tasks like processing audio, text, and images together in healthcare diagnostics or creative media. At Wipro, the choice between SLM or LLM is based entirely on client requirements and the scope of the project.
When asked if AI misuse be prevented, especially by people with malicious intent?
Rajeev explained that modern AI systems have three protective layers: the user interface, the data processing layer, and the guardrail layer.
The guardrail monitors queries to detect and block inappropriate or unethical ones. Moreover, usage patterns are constantly tracked, and feedback from these interactions is used to strengthen the model over time. This includes not only security but also improving response quality.
Tell us a little about your personal journey into AI. Was it always part of your plan?
Rajeev shared that his journey into AI began by chance. Two years ago, AI was still emerging and most people were chasing more established tech roles. But he saw an opening and decided to take a leap. A turning point came during his MBA at IIM Kozhikode, when Professor Raju told him, 'The next decade belongs to data. If you control data, you control the world.' That advice inspired him to pivot, and it turned out to be a defining decision in his career.
Was there a specific moment that confirmed you made the right choice?
Yes, Rajeev recalled a friend who struggled to manually sift through 1,000 job applications. In just five days, Rajeev built a tool that could score resumes against job descriptions. To make it more robust, he implemented cosine similarity to detect AI-generated or overly similar resumes, helping to remove redundant applications. That moment made him realize how practically powerful and impactful AI can be.
AI-generated resumes are becoming common. Could the best candidates be overlooked?
Rajeev said it's a real concern. Many candidates now tailor their resumes to pass AI filters using keywords and tools. While this helps visibility, it also leads to over-standardization, which might mask real talent. He advised applicants to be strategic, 'Use AI to enhance your resume, but remember that authentic skills and substance still matter most.'
Rajeev concluded that AI is a tool, not a threat. If used responsibly, it has the potential to enhance human capabilities, not replace them. The key is to stay ethical, curious, and collaborative. "AI is here to stay. The question is: how responsibly and creatively will we use it?"
Interview by: Gyanisha Mallick
Guest: Rajeev Ranjan Kumar, Senior Leader & AI specialist, Wipro
Platform: The Hans India; TechTalk Podcast