GCCs shifting from outsourcing, they're now part of core product teams: SNDK Corp's Founder-CTO

As Global Capability Centers (GCCs) in India move beyond outsourced execution to focus on R&D and emerging technologies, companies are rethinking how they build and scale tech teams.

In a conversation with TechCircle, Brijesh Patel, Founder and CTO of SNDK Corp, a tech consulting firm based in Ahmedabad, talks about what this shift looks like on the ground. Working with clients in manufacturing and Fast-Moving Consumer Goods (FMCG), Patel shares how businesses are approaching Artificial Intelligence (AI), automation, and blockchain, not as buzzwords, but as tools to solve specific operational problems. From AI-driven support systems to predictive maintenance, this conversation looks at how emerging tech is being applied in practical, measurable ways. Edited Excerpts:

How have you seen the role of GCCs in India evolve over the years, and where do you think they're headed next?

People have started recognising the value of technical and research talent in India. Earlier, it was mostly about getting execution done, developers were seen as just resources to complete tasks. Now, there’s more investment in R&D.

Companies are looking to build smaller, more efficient teams. Hiring an AI expert in the US is expensive, so they’re setting up counterparts in India. The language and communication barriers are decreasing, and with strong local management, teams here can support high-level R&D in areas like AI, analytics, cloud computing, and blockchain.

This isn’t a plug-and-play process. It takes time to identify internal use cases and integrate these technologies. As a result, companies are moving away from outsourcing everything to third-party vendors. Instead, they’re building their own centers in India, directly connected to their US stakeholders, as I did for a US-based food company. This creates a more integrated and aligned ecosystem.

Setting up in India has also become easier. The government has helped reduce bureaucratic hurdles, making it simpler for global companies to tap into the local tech talent.

You work with SMEs in manufacturing and FMCG, how do their AI needs differ across industries, and what challenges do they face in adopting AI?

The biggest challenge for most SMEs, whether in the US or India, is cost. Return of Investment (RoI) is critical, so it's important to identify the right use case, typically a process that relies heavily on human effort and offers a clear return.

When we started exploring AI for a food tech company, the goal was to build a chatbot that could query data and return answers in natural language. The User Interface (UI) and analytics were already in place, but we wanted marketing teams to interact with the data more effectively via chat.

Initially, setting up the required infrastructure, servers and other components, was too expensive. Six to eight months later, things changed. Platforms like AWS Bedrock made it possible to deploy similar solutions more affordably, on a pay-as-you-go basis.

We launched our solution on AWS using Anthropic on the backend, though you can use any model that fits your needs. Now, creating AI POCs or case studies is far more accessible. With lower costs and more available tools, AI adoption is accelerating. New developments are also emerging constantly, making this a fast-moving space.

Most companies use AI mainly for chatbots. Your team has applied it in areas like HR and SOP automation. What's one AI use case that unexpectedly delivered strong business value?

I break this down into three areas: time, resources, and energy. If I can help businesses reduce the time it takes to complete certain tasks, the process can eventually be integrated into their broader business workflows, which leads to long-term cost savings.

For example, we recently built an open-source VoIP phone system for a global company. They use our local support center to handle their L1, L2, and L3 technical support.

We started recording all incoming support tickets and automatically converting them into entries in our custom CRM. These tickets are summarised using AI and include the full conversation between the support rep and the user.

Each ticket is automatically categorized, CRM issues, ERP problems, network issues, printer problems, etc. AI handles this tagging and improves over time. The next step is agentic AI, where we replace the initial human interaction with an AI bot. This bot will ask users standard diagnostic questions and then create and route the ticket to the right person, just as a human would.

As a result of implementing this system, support ticket efficiency improved by nearly 60%, with gains in RoI, accuracy, and accountability.

How do you handle data localisation in your SaaS or AI products for global vs. Indian clients?

We’ve been encouraging our clients to move to cloud platforms like AWS, which are already compliant with standards such as SOC, GDPR, and HIPAA. This takes care of a major portion of security requirements.

If we were to handle everything on-premises, the cost and complexity would be much higher. That said, it’s still possible. We also conduct regular audits, both internal and third-party, and have a cybersecurity team that oversees all security aspects.

We follow industry-standard best practices to manage data security. Security is always a priority, regardless of the project size. Even with something as simple as a website, we ensure it’s secure and that client data is protected, encrypted, and properly stored.

For high-priority or compliance-heavy projects, we also work with third-party auditors and partner companies to ensure all compliance requirements are met.

Given the increasing cybersecurity threats, especially over the past six months, this focus is more critical than ever. We’ve also started using AI tools like Cursor and Amazon CodeGuru for code reviews. These tools help us identify security issues, such as hardcoded credentials or other vulnerabilities, and have proven to be very effective.

How do you see India’s growing AI infrastructure, like GPU clusters and government-backed platforms, impacting the SME tech ecosystem and your roadmap?

At some point, we will need to develop our own AI model specifically for India. The country’s diversity and complexity mean that a one-size-fits-all approach won’t work. India has its own languages, cultures, user behaviors, regulatory needs, and technical ecosystem. These factors require an AI system that understands and adapts to local realities.

To achieve this, we’ll need a coordinated effort involving the government, technology companies, academic institutions, and everyday users. Everyone will need to contribute to a shared platform that supports the development of AI tools built on top of a common infrastructure. This platform should enable innovation while ensuring that solutions are affordable, secure, and aligned with Indian norms and requirements.

Right now, most widely used AI models are developed in the US or other countries, based on their data, languages, and user contexts. While these models are powerful, they don’t fully meet India’s needs. Building our own model, trained on Indian languages, datasets, and cultural contexts, would allow us to create solutions that are more relevant, more accessible, and more effective for local use.

In the long run, this would not only help address domestic challenges but also position India as a serious player in the global AI landscape, with expertise rooted in solving problems at scale in a complex environment.

Looking ahead, which emerging technologies, like blockchain or predictive AI, are you most interested in, and how do you plan to use them in your work?

One of our long-term goals has been to implement blockchain. Right now, we're working on integrating blockchain into our newer ERP and CRM systems, including the communication platforms we've built. We're exploring the best way to connect blockchain with these systems to make everything traceable.

The goal is to ensure that even if our platform connects with external systems, we maintain traceability and security. Alongside this, we're also focusing on predictive maintenance, as our background is in food manufacturing and distribution.

We're partnering with Ignition through Inductive Automation, a major player in Industry 4.0. They're well-established in the US and other countries, but adoption in India is still limited, mostly among top-tier companies, due to the high cost and the investment required to get meaningful machine insights.

Despite that, we see strong potential. We're currently collecting machine data over a period of one to three years to build and train custom AI models. The idea is to use historical data to detect patterns that have previously led to machine failures. When similar patterns emerge, the system will alert the relevant personnel to perform maintenance.

This project aims to create predictive models for each type of machine so that, within a year or so, companies can use their own data to anticipate issues and reduce downtime.

---