How does Newgen Software handle multilingual document processing across India’s states, where citizen documents may be in regional languages with inconsistent formatting or handwriting?
India is probably one of the best examples of why intelligent document processing cannot be solved with a simple OCR engine or a collection of fixed templates. We have enormous diversity in languages, scripts, document formats, and information quality. Add handwriting, scanned documents, and documents from different government departments, and the problem becomes one of understanding context rather than just reading text.
AI Today is the game changer. The technology must be able to identify what a document is, understand the information within it, and determine what is relevant to the subsequent process. It also must know when it is not confident enough to make a decision. We do not believe that everything should be automated simply because it can be. In government, accuracy and accountability matter as much as speed.
Newgen solves this complicated problem through a mix of technologies, including AI, OCR, OMR, and Barcode technologies. We break the form down into individual zones and apply the right technology for the best results. Our approach is therefore to combine document intelligence with the workflow around it. The real value is not extracting information from a document. It is taking that information and moving the citizen’s application, claim, request, or case forward without unnecessary manual intervention.
That is the larger opportunity in India. We can use AI to absorb the complexity of our languages and documents while making the experience at the other end much simpler for the citizen.
Tarun Nandwani, CEO, Newgen Software
With 67 patents filed and 25 granted, which areas of Newgen Software’s IP portfolio does the company consider most defensible against larger global competitors?
I would not define our moat by the number of patents we have. Patents are important, but they are only one part of defensibility.
What is much harder to replicate is the combination of technology, domain knowledge and the experience of taking that technology into some of the most complex enterprise environments.
We have spent more than three decades working with processes where content, decisions, people and systems all have to work together. That experience has shaped the platform we have today.
The first area of strength is our understanding of enterprise content. Enterprises do not operate on clean, structured data alone. A huge part of business knowledge still sits inside documents, correspondence, records, and other unstructured content. Being able to understand that content and bring it into a business process is a significant capability.
The second is the convergence of AI with process automation. We are not trying to build AI as a standalone capability. We want intelligence to sit inside the processes where decisions are actually made, and work gets done.
And increasingly, the third area is trusted AI. Enterprises, particularly in banking, insurance, and government, are not looking for an AI system that simply gives them an answer. They need to know where the answer came from, what information was used, what rules were applied, and who remains accountable for the outcome.
That combination is difficult to replicate. A competitor can build a feature. It is much harder to replicate decades of enterprise experience, a mature platform, domain knowledge, and the ability to operationalize AI in mission-critical environments.
How does Newgen Software work with government departments that have deeply entrenched legacy systems already handling live citizen data, without disrupting continuity of service during migration?
The biggest mistake in modernization is to think that you have to replace everything before you can improve anything.
Government systems differ from many enterprise systems because they deliver live citizen services. You cannot simply stop a system because a new one is ready. The cost of disruption is ultimately paid by the citizen.
Our approach is therefore evolutionary rather than disruptive. We start by looking at where technology can create value immediately without disrupting the systems that are still doing important work. That could mean connecting existing systems, improving a workflow, bringing intelligence to document processing, or creating a better experience around an existing system of record. Then, as the organization gains confidence, individual capabilities can be progressively modernized and migrated.
This is also why we believe an orchestration layer is so important. It allows organizations to connect people, processes, content, data, and existing systems while they modernize underneath. You do not have to make modernization an all-or-nothing decision.
We have seen this approach work in large government environments where significant volumes of information and multiple legacy platforms have to be brought together. The philosophy is simple: do not modernize for the sake of modernizing. Modernize where it creates measurable value, protect what is working, and gradually retire what is no longer needed.
For the government, successful modernization is not about having the newest technology. It is about improving the service without jeopardizing continuity.
What safeguards does Newgen Software build into AI-powered public sector workflows to prevent errors in automated decision-making that could affect a citizen’s access to benefits or services?
This is where I think the conversation around enterprise AI needs to become more mature. There is a big difference between using AI to assist a government employee and asking AI to make a decision that affects a citizen’s rights or access to a service. We should not treat those two use cases in the same way. The more consequential the decision, the more important governance becomes.
For us, AI has to operate within the context of the organization’s own information, policies, and processes. It should not be making decisions based on an isolated prompt or information that nobody can validate.
It also needs boundaries. An AI agent should know what it is allowed to do, what it is not allowed to do, and when a human needs to take over. That means business rules, permissions, escalation mechanisms, and auditability have to be part of the architecture, not something added after deployment.
And finally, there has to be accountability. If an AI system recommends an action, particularly in a sensitive government process, there must be a way to understand the basis of that recommendation and review or override it when necessary.
I believe this is going to become one of the defining questions of enterprise AI. The winners will not simply be the organisations that deploy the most AI. They will be the ones that can deploy it at scale while maintaining trust.
Responsible AI, therefore, is not a policy document for us. It has to be built into the way the technology operates.
What role does cloud versus on-premises deployment play in Newgen Software’s national infrastructure projects, especially in departments handling sensitive citizen data?
I do not think the right question is whether government should choose cloud or on premises. The right question is what the nature of the workload demands.
Government has to deal with data sovereignty, security, resilience, regulatory requirements and continuity of essential services. Those considerations will be different for different workloads and different departments.
For highly sensitive workloads, there may be a strong case for keeping data and processing within tightly controlled environments. For other workloads, cloud can provide significant advantages in scalability, resilience, speed of deployment and the ability to innovate faster.
So, I expect the future to be increasingly hybrid.
What matters is that the technology platform should not force the government into one infrastructure choice. It should allow the organisation to modernize at its own pace and choose the right deployment model based on the sensitivity and business criticality of each workload.
There is also a broader point here. Cloud adoption should not become the modernization strategy by itself. The objective is better government services, faster decision-making and more efficient operations. Infrastructure is an enabler of that transformation. Our role is to give governments that flexibility while maintaining the security, control and continuity that public infrastructure demands.
Where does Newgen Software think Indian e-governance will be in five years if the “Intelligent Enterprise” model is adopted at scale, versus if it isn’t?
I think the next five years will determine whether we simply make government digital or actually make it intelligent. India has already done an enormous amount of work in digitizing services. The next opportunity is to make those services work together much more intelligently.
Today, a citizen can still experience government as a collection of departments, forms, portals, and processes. The future should be very different. The government should increasingly be able to understand the context of a citizen’s interaction, use existing information, connect relevant departments, and automate routine parts of the journey.
The citizen should not have to understand the structure of government to obtain a government service. AI can be a major catalyst for this shift, but AI alone is not the answer. You need the underlying processes, content, data, and systems to be connected. You need governance. You also need a strong foundation of data security, privacy, and access controls so that enterprise AI can operate responsibly while maintaining citizen trust and protecting sensitive public information. And you need the ability to orchestrate all these pieces around an outcome.
If we get that right, I see Indian e-governance becoming much more proactive. Services can move from being something a citizen has to initiate and navigate to something that increasingly anticipates needs and resolves them with minimal friction.
If we do not make that transition, we risk creating a digital version of the same fragmentation we have today: more portals, more applications, and more technology, but the same complexity underneath.
For me, the real definition of an intelligent government is quite simple. The citizen should not have to think about the technology behind the service. The technology should simply make the service faster, simpler, and more responsive. That is where I believe India has the opportunity to lead.
About The Author
