In this exclusive IT Voice interview, Niraj Kumar, CTO of Onix, shares deep insights into how AI is reshaping business operations, tackling data biases, and accelerating cloud-to-AI transitions through their groundbreaking Wingspan platform. Discover how Onix is enabling responsible, scalable, and context-aware AI adoption for the enterprise future.
1. AI-Powered Business Operations
IT Voice- How do you see AI transforming business operations in 2025 and beyond?
Niraj Kumar: AI is no longer just a futuristic concept. It has become an essential tool for solving real-world business challenges. In 2025, AI’s evolution will be marked by practical, secure, and transparent implementations, shifting from broad applications to highly specialized, industry-specific use cases. This shift, powered by Generative AI and multimodal capabilities, will redefine operations in key sectors like healthcare, finance, and supply chain management, enabling more precise decision-making and streamlined workflows.
One of the most significant transformations will be augmented intelligence, where AI and human expertise work in tandem to enhance decision-making. By automating routine tasks and delivering real-time, actionable insights, AI will free up professionals to focus on creativity, strategic thinking, and innovation.
Furthermore, the convergence of data, Generative AI, and search technologies will reshape how businesses extract value from their data. This powerful combination will drive hyper-personalization, smarter decision-making, precise customer targeting, and the rapid discovery of new opportunities. As Enterprise Search and AI-driven analytics mature, organizations will gain unprecedented agility, unlocking new efficiencies and competitive advantages.
AI will no longer be an optional enhancement, it will be the foundation of business transformation, driving efficiency, innovation, and growth in ways that were previously unimaginable.
2. Addressing AI Biases and Data Governance
IT Voice- Despite advancements, AI still struggles with bias. What strategies do you recommend for mitigating these biases?
Niraj Kumar: Bias in AI remains a significant challenge, especially as these models become more embedded in decision-making processes. At Onix, we believe that mitigating AI bias requires a multi-pronged strategy. One of the most effective approaches we advocate is the use of synthetic data. It allows us to create balanced, representative datasets that overcome historical bias without compromising privacy.
We also emphasize building explainability into our models, ensuring that teams understand how decisions are made and can intervene when necessary. Regular audits, diverse training datasets, and human-in-the-loop frameworks are equally critical. Ultimately, responsible AI is not just about technology, it’s about governance, ethics, and cross-functional collaboration. And as marketers and communicators, we have a role to play in making these conversations more accessible and transparent.
3. Cloud and Data Strategy Evolution
IT Voice- How does Wingspan accelerate the transition from data to cloud to AI, ensuring enterprises unlock value faster from their cloud and AI investments?
Niraj Kumar: Wingspan accelerates the transition from data to cloud to AI by offering a multi-capability agentic AI platform that seamlessly integrates Onix’s proprietary data modernization technologies with context-aware AI agents. This integration allows enterprises to streamline their data-to-AI transformation, ensuring faster realization of value from their cloud and AI investments. By automating critical data processes and enabling the rapid deployment of domain-specific AI solutions, Wingspan helps businesses unlock actionable insights and drive efficiency.
A key advantage of Wingspan is its ability to accelerate AI and data platform adoption by 2-3 times compared to traditional methods. Enterprises leveraging Wingspan can bring AI initiatives into production within just four weeks, significantly reducing the time required to move from data modernization to AI-driven decision-making. Its seamless integration with Google Cloud AI products, including Google Agentspace, simplifies AI deployment, making it easier for organizations to harness the full potential of cloud and AI without operational bottlenecks.
IT Voice- Wingspan is the first multi-model agentic AI platform in the industry. What challenges or gaps led to its development?
Niraj Kumar: Wingspan was developed to address key challenges in enterprise AI adoption, including slow deployment, fragmented data ecosystems, and a lack of contextual intelligence in AI-driven decision-making. Traditional AI solutions often require extensive customization, take months to move from proof of concept to production, and struggle to integrate seamlessly with existing business processes.
One major gap was the complexity of AI deployment—enterprises needed a solution that could unify data, lineage, and AI while ensuring scalability. Existing AI models also lacked context-awareness, making it difficult to generate meaningful, business-specific insights. Additionally, most AI solutions require constant human intervention, limiting automation and slowing adoption.
Wingspan, powered by multiple AI agents, fills these gaps with autonomous and deterministic AI agents capable of operating within an enterprise’s specific business language and processes. By leveraging Onix’s proprietary context engine IPs and integrating with Google Cloud AI products, including Google Agentspace, Wingspan enables enterprises to bring AI solutions into production within just four weeks. This industry-first platform accelerates data-to-AI transformation, ensuring businesses can unlock value faster, drive smarter decisions, and scale AI adoption efficiently.
4. The Ethical AI Imperative
IT Voice- Responsible AI adoption is a growing concern. What measures should companies take to ensure transparency and fairness in AI?
Niraj Kumar: Responsible AI is no longer optional; it’s a business imperative. To ensure transparency and fairness, companies need to embed responsibility at every stage of the AI lifecycle. At Onix, we focus on a few key principles. First, transparency starts with explainability. Whether it’s a model used in HR, healthcare, or finance, stakeholders must understand how and why decisions are made.
Second, fairness comes from inclusive data and diverse development teams. We advocate for bias detection tools, regular model audits, and the integration of synthetic data to correct imbalances in training datasets.
Third, governance is essential. Clear guidelines, accountability structures, and cross-functional AI ethics boards help ensure that AI systems are aligned with organizational values and societal norms.
Ultimately, responsible AI is about creating trust, not just in the technology but in how it’s developed, deployed, and communicated.
IT Voice- Many enterprises hesitate to embrace full automation due to governance, compliance, and data integrity concerns. How does Wingspan’s ‘human-in-the-loop’ approach strike the right balance between autonomy and control?
Niraj Kumar: That hesitation is completely valid—governance, compliance, and data integrity are foundational concerns, especially in highly regulated industries. Wingspan was built with that reality in mind. Our ‘human-in-the-loop’ approach isn’t just a safeguard, it’s a design philosophy that strikes the right balance between automation and oversight.
By enabling humans to validate, override, or retrain AI outputs at critical checkpoints, Wingspan ensures that enterprises maintain control without sacrificing the speed and efficiency of automation. This approach helps organizations scale AI responsibly, with audit trails, policy enforcement, and transparency built into the process.
In essence, Wingspan empowers teams to trust automation not by removing the human element but by making it smarter and more context-aware. That’s how we deliver both innovation and assurance.
5. AI-Driven Customer Experiences
IT Voice- Context-aware AI is a major breakthrough in enterprise technology. Can you explain how Wingspan understands business context and adapts to unique modernization needs without human intervention?
Niraj Kumar: Context-aware AI is a game-changer for enterprise transformation, and it’s one of Wingspan’s most powerful capabilities. What makes it unique is its ability to understand not just data, but the meaning behind it, how it connects to business processes, domain-specific language, and operational goals.
Wingspan achieves this through our enterprise knowledge graph, Eagle, which maps the organization’s data ecosystem and business logic. Combined with proprietary context engines like Raven, Pelican, and Kingfisher, Wingspan enables AI agents to operate with deterministic precision, learning, adapting, and taking action without constant human intervention.
This means AI becomes less about managing models and more about driving outcomes. And that’s exactly why we built Wingspan, to make AI adoption efficient, effective, and genuinely enterprise-ready.
6. Data as a Product
IT Voice- The concept of “Data-as-a-Product” is gaining traction. How can organizations monetize their data effectively?
Niraj Kumar: To effectively monetize data, organizations must shift towards a “Data-as-a-Product” approach, treating data as a strategic, high-value asset rather than a byproduct of operations. This requires ensuring that data is complete, clean, and contextually enriched, making it ready for use across various business functions and AI-driven applications.
A key enabler of data monetization is data augmentation and synthetic data technologies, which help address data scarcity and privacy concerns. By generating high-quality synthetic data, businesses can create valuable datasets without compromising sensitive information, opening new revenue streams through data-sharing partnerships, AI model training, and industry collaborations.
Additionally, real-time analytics and context-aware systems allow organizations to anticipate trends, enhance decision-making, and extract deeper insights from their data assets. Tools like Eagle play a crucial role in this transformation by helping businesses understand their data ecosystems, identify gaps, and turn fragmented datasets into actionable intelligence.
By adopting a data-first strategy, enterprises can unlock new growth opportunities, drive AI innovation, and create scalable data products that fuel competitive advantage and long-term business success.
IT Voice- What are the key risks and challenges associated with data commercialization?
Niraj Kumar: Data commercialization holds tremendous potential, but it also brings several risks and challenges. Privacy and security concerns are at the forefront, especially with strict regulations like GDPR. There’s also the issue of data quality, poor or biased data can lead to faulty insights. Additionally, ethical concerns around data misuse or discrimination are crucial to address. Finally, striking the right balance between monetizing data and maintaining customer trust is a constant challenge for enterprises venturing into data commercialization.
7. Synthetic Data Generation
IT Voice- What are the major advantages of using synthetic data over real-world datasets?
Niraj Kumar: Synthetic data offers several key advantages over real-world datasets, making it a powerful tool for AI development, analytics, and business decision-making. One major benefit is privacy and security. Since synthetic data is artificially generated, it does not contain personally identifiable information (PII) or sensitive business data, reducing compliance risks and ensuring adherence to strict data privacy regulations like GDPR and HIPAA.
Another advantage is addressing data scarcity and bias. Many industries struggle with limited access to high-quality real-world data, particularly in areas like healthcare and finance. Synthetic data enables organizations to generate diverse, balanced datasets, reducing bias in AI models and improving overall accuracy. Cost efficiency and scalability are also key benefits. Collecting and annotating real-world data is expensive and time-consuming, whereas synthetic data can be created quickly and at scale, enabling rapid AI model training and iteration.
Additionally, synthetic data allows for enhanced control and customization. Businesses can simulate rare scenarios, edge cases, and specific conditions that may be difficult to capture in real-world datasets, improving AI robustness and predictive capabilities. By leveraging synthetic data, organizations can accelerate AI development, reduce risks, and drive innovation while ensuring compliance and cost-effectiveness.
IT Voice- Can synthetic data help bridge gaps in AI training while addressing privacy concerns?
Niraj Kumar: Yes, synthetic data is a powerful tool to bridge gaps in AI training, especially in areas where real-world data is limited, sensitive, or imbalanced. It allows organizations to generate high-quality, anonymized datasets that not only protect individual privacy but also improve model performance by covering edge cases and reducing bias.
At Onix, we’re taking this a step further with Kingfisher, our synthetic data generation engine that works in tandem with Wingspan. Kingfisher helps enterprises simulate real-world scenarios at scale while maintaining strict compliance and ethical standards. It’s particularly valuable in sectors like finance and healthcare, where data privacy is non-negotiable. With Kingfisher, we’re enabling AI development that’s both powerful and principled.
8. Preparing for AI-First Workplaces
IT Voice- What advice would you give to business leaders looking to integrate AI seamlessly into their operations?
Niraj Kumar: My advice to business leaders is to start with purpose, not just technology. Successful AI integration begins by identifying real business problems and matching them with practical, measurable solutions. Start small, test, and scale what works. And just as importantly, ensure that AI systems are explainable, auditable, and embedded with governance from day one.
At Onix, we realized that while the ‘data to AI’ journey sounds simple, the reality is full of fragmentation and hidden gaps that derail transformation. That’s exactly why we created and recently launched Wingspan, to help businesses integrate AI seamlessly into their operations, with clarity, control, and confidence.
