Amit Pradhan on scaling AI from pilot to production, building digital manufacturing capabilities, and preparing Indian enterprises for the future of autonomous operations.
One of the biggest challenges enterprises face today is not adopting AI—but scaling it. As companies move from small pilot projects to enterprise-wide AI adoption, they face a host of challenges, including data readiness, system integration, change management, and aligning with business goals. Amit Pradhan, Vice President – IT and Chief Information Officer (CIO) at Dixon Technologies (India) Ltd., has been leading this transformation from the front.
Dixon Technologies is one of India’s leading electronics manufacturing services (EMS) companies, partnering with top global and Indian brands across segments such as consumer electronics, home appliances, mobile phones, lighting, and wearables. As a key contributor to the “Make in India” initiative, Dixon is actively building digital capabilities to scale smart manufacturing and foster innovation.
Amit brings over 20 years of cross-industry experience in electronics, telecom, energy, and consumer durables. At Dixon, he heads the company’s IT strategy, digital initiatives, and innovation-led transformation agenda. Prior to joining Dixon, he held senior leadership roles at Mahindra Group, Sterlite Power, Adani Enterprises, and Videocon, where he drove several large-scale technology initiatives. A NEXT100 awardee (2019), Amit combines deep enterprise leadership with entrepreneurial agility.
In this exclusive conversation with Jatinder Singh, Executive Editor, CIO&Leader, and Vikas Gupta, Editorial Director, 9.9 Group, Amit shares how Dixon is scaling AI, strengthening digital manufacturing, and preparing for the future of autonomous enterprise operations.
What are some of the unique challenges that companies like yours face in operationalizing AI? And what best practices have emerged from your experience?
AI as a concept is not new. What has changed is the maturity of the ecosystem, especially for enterprise-scale deployment. A few years ago, AI was limited to isolated use cases, often inspired by Big Tech, but rarely scalable across traditional industries like manufacturing. The rapid pace of change and frequent model updates made it difficult to stabilize.
Today, the ecosystem is far more mature. At Dixon, we have built and tested nearly 20 AI proof-of-concepts (POCs). Of these, 5–6 are now being scaled across the organization including globally.
Being a manufacturing-intensive company, our AI focus is largely on production, supply chain, and testing. Some key examples include:
- Dixon answer platform: We built an internal platform layered over our DMS, CMS, and other systems to create a bot that serves as a central knowledge assistant. Employees no longer need to sift through long documents—they simply ask the bot and receive precise, contextual answers on policies, quality checks, or procedures.
- Defect detection and prevention: We use AI models to detect product defects and recommend corrective actions. This enables continuous learning and prevention, improving product quality.
- Supply chain optimization: One of our key projects is an AI-powered supply planning tool. It identifies alternative components when existing ones reach end-of-life, helping us manage inventory efficiently and support tech refresh cycles with minimal waste.
Laptop and microphone manufacturing are relatively new verticals for us, and we face a shortage of skilled manpower—especially for operating SMT machines, just to give a precise example. This makes it essential for us to focus on both scaling operations and effective manpower mapping.
To address this, we are using AI-driven video-based training solutions. Traditional classroom sessions or lengthy training videos are often ineffective in such environments. So, we have created short, targeted training clips—like step-by-step guides—that help operators quickly learn how to handle specific tasks or machines. These are focused on ensuring quality and efficiency on the production floor.
How are you approaching data governance to ensure AI implementations are effective and sustainable?
AI is only as good as the data that feeds it. That’s why data governance is a foundational part of our AI strategy. We have transitioned from Excel-based planning to an integrated IT platform, which ensures master data is accurate and consistently maintained—especially important for our supply chain planning.
We have also implemented a maker-checker model that started manually but has now evolved into an intelligent system. It automatically flags anomalies and discrepancies, allowing us to act before issues escalate.
A key goal has been to establish a single source of truth, which is critical in a complex setup with multiple data streams. This ensures reliability and trust in our AI outputs and supports better decision-making.
What are your key priorities for AI and technology in the next couple of years?
We are focused on three strategic priorities:
- AI strategy – maker, saver, taker:
Ø Maker: Build custom AI models for our unique business needs.
Ø Saver: Leverage and refine existing internal models.
Ø Taker: Integrate SaaS platforms with pre-built AI capabilities.
- Edge-to-cloud strategy:
- Our fast-paced production lines require minimal latency. We are exploring Edge-to-Cloud architectures that combine local responsiveness with cloud scalability.
- Transitioning from Industry 4.0 to 5.0:
- We are deploying technologies like intelligent UIs, advanced automation, camera-based quality checks, and AI-guided workstations.
- A strong focus is on workforce skilling—we use AI-driven video tutorials and short, gamified training clips that are process-specific and sometimes even certified. In industries like ours, where skilled manpower (e.g., SMT operators) is limited, this makes a big difference.
- We’ve also introduced digital displays at workstations with real-time, interactive instructions—significantly improving both quality and productivity.
Is there growing pressure from global brand partners to include AI in your manufacturing lines, or are these initiatives led purely by Dixon?
It’s a mix of both. On one hand, we are internally driving AI as part of our KPIs—to make manufacturing more efficient, sustainable, and smarter.
On the other hand, global customers are increasingly expecting their partners to be tech-enabled. One of our largest clients recently adopted an AI-powered logistics quality solution we developed and told us, “This is the first time any supplier worldwide has done this.” They’re now looking to roll it out with other vendors too. That’s a big validation of our efforts.
From your vantage point, do you think fully autonomous enterprises will be a reality in India within the next two years?
I believe we are moving in the right direction. Many Indian enterprises have already set up dedicated AI teams and are integrating AI across departments.
Of course, for some companies, it’s still a journey—especially where digital maturity or foundational data infrastructure is lacking. But the momentum is strong. For enterprises that are still building AI foundations—like working on surrounding technologies, developing new skills, and scaling data infrastructure—it’s definitely a journey. It will take a bit more time for them to fully adopt and integrate AI. But as I said, AI is a powerful game-changer. In just a couple of years, I believe around 80% of our work will be either AI-assisted or entirely bot-driven.
What kind of topics would you personally like to see covered at CIO&Leader’s upcoming conference on “AI: From Pilot to Production”?
There are three areas I’d like to learn more about, especially given the challenges CIOs are currently facing:
- Agent-based AI:
There is significant promise here—where AI agents can connect multiple systems, data sources, and actions to replicate decision-making. - Responsible AI:
As more models are deployed, trust, bias mitigation, and transparency become critical. We need frameworks to ensure ethical and accountable AI use. - India-specific AI innovations:
I would love to see how India-built AI stacks and where they can be practically applied. Understanding local innovation will help contextualize global trends.