Simplicity, speed & scalability are the key pillars of our AI strategy: Siddharth Sureka, Motilal Oswal Financial Services

The BFSI sector has rapidly evolved from viewing AI as an experimental tool to a core strategic asset. How would you describe Motilal Oswal’s journey in this landscape, moving from initial pilot use cases to enterprise-scale implementation?
You have rightly identified the trajectory. If we look back, 2023 was truly the year of experimentation for the industry, followed by 2024, which was defined by pilot use cases. Now, as we move through 2025, AI has definitively arrived in the enterprise. At Motilal Oswal, we have taken a very long-term view on how to embed AI into the very fabric of our organisation because we recognise it as a fundamentally transformative technology.

AI is here to stay, and will transform all industries. Naturally, the BFSI sector tends to be on the leading edge of this journey, following closely behind pure technology companies. However, rather than viewing this purely through a technology lens, we approached it from an end-to-end organisational transformation lens. The core question we asked ourselves was – how do we re-imagine and embed this power across the entire value chain?

To achieve this, we prioritised identifying use cases that offered measurable business outcomes and could be delivered promptly. This strategy allowed us to build the necessary momentum to tackle the larger, more complex opportunities that lay ahead. We moved from a traditional Software Development Life Cycle (SDLC) to what we now call the AI Development Life Cycle (AIDLC). This shift has allowed us to shrink the time from ideation to development significantly. This means what previously took us three months to build, we are now able to execute in two months. This represents a 33% efficiency impact, which ultimately benefits our clients, who are at the centre of everything we do.

Given Motilal Oswal’s diverse portfolio, spanning broking, distribution, and asset management, what is the overarching AI strategy guiding these initiatives? How do you ensure these efforts drive tangible outcomes like financial inclusion or AUM growth?
When I was on-boarded about a year and a half ago, following my tenure leading cognitive AI at Charles Schwab, the management gave me a very clear brief – how can we democratise finance in India? The goal was to increase penetration and adapt to a new world where AI is at the centre. To achieve this and create competitive differentiation, we broke our strategy down into three core pillars i.e. simplicity, speed, and scale.

The first pillar is simplicity. To reach tier two, three, and four cities, we must make the financial experience intuitive. Simplicity is driven by personalisation, which means how we curate the information delivered to clients and ensure their digital journey is frictionless.

The second pillar is speed. We are in the business of providing the right insights at the speed of the market. As an event occurs, we must be able to serve our clients with immediate insights. A prime example of this is our ‘News Agent’ product. As news arrives, the system measures the sentiment and analyses how it may impact the market, and then serves that insight directly to the client instantly.

The third vertical is scalability. Once we have achieved simplicity and speed, our focus is to scale this architecture to reach the deeper pockets of the country. This scalability is essential for the financial inclusion journey we are embarked upon, ensuring that investors in tier three and four cities can take full advantage of the markets.

How do you measure the ROI of your AI investments to ensure they are delivering true organisational value?
This is an extremely important point. A recent report suggested that nearly 90% of AI investments fail to yield the expected ROI. And I often joke that AI is everywhere, but not a single drop of ROI is to be found. The primary reason organisations fail to see ROI is that they treat AI projects like standard software engineering projects.

In software engineering, you are delivering a deterministic output. However, when you move into the domain of AI, the outcomes become stochastic or probabilistic in nature. As leaders, we must understand the use cases we are working on and, crucially, the ‘cost of getting it wrong’.

We measure our success by tracking efficiency and output very deliberately. For example, in our software engineering use cases (AIDLC), we measured the ‘story points’ delivered in a sprint before and after AI adoption, noting a distinct increase in efficiency. Similarly, in our marketing function, where we generate over 500 creatives a month across our group companies, the volume and time-to-market improvements were clearly measurable.

However, because these are probabilistic systems, we always maintain a ‘human in the loop’ structure. Whether it is a technical engineer reviewing code or a marketing professional reviewing a GenAI-generated creative, human oversight ensures accuracy. If the system gets it wrong, which it might, as it is not 100% deterministic, the human expert corrects it. This structure allows us to capture the efficiency gains of AI while mitigating the risks inherent in probabilistic models.

Beyond operational efficiency, how is Motilal Oswal leveraging GenAI to innovate the customer experience, specifically regarding personalised advice or research consumption? What unique challenges does a highly regulated sector like financial services pose for GenAI deployment?
We are moving beyond simple chatbots to truly transformative customer engagement. Our ‘North Star’ remains the democratisation of finance, and generative AI is a key lever in moving that needle forward.

We are currently preparing to launch a ‘Research Assistant’ or Research GPT. At Motilal Oswal, we pride ourselves on the depth of our research, covering about 300 companies. However, institutional research reports can be 30 pages long and complex, making them difficult for an ordinary retail investor to consume.

GenAI plays a huge role here by synthesising these complex pieces of information. The Research Assistant allows a user to simply ask, ‘Give me a view on Motilal Oswal’. The system will analyse our proprietary research and industry data to provide the company’s performance, our recommendation, targets, and stop-loss details in an easily digestible format. This empowers retail investors with decision-making capabilities that were previously not at their fingertips.

Regarding the challenges in a regulated sector, we manage this by categorising the cost of error. If the cost of getting it wrong is high, such as delivering impactful financial research, we enforce strict human oversight. If the cost is low, such as a slight personalsation error in a marketing image, we have more latitude. This tiered approach to risk allows us to deploy GenAI successfully without compromising our fiduciary duties.

High-quality data is the fuel for reliable AI. Could you detail your strategy for data governance and the technology stack you are prioritising to ensure secure, unbiased, and scalable AI deployment?
Data governance is top of mind because, with stochastic systems, the principle of ‘garbage in, garbage out’ applies strictly. Motilal Oswal began its data journey around 2020, establishing a Data Lake as the foundation of our architecture.

To manage this effectively, we adhere to a Responsible AI Framework, which focuses on several key areas:

• Data provenance and lineage: Before any data enters the model building process, we must understand its origin, its lineage, and run Automated Quality Checks (AQCs) to identify and mitigate biases.

• Transparency and explainability: We use techniques like SHAP (SHapley Additive exPlanations) values to understand model outcomes. It is vital that we can explain why a model made a specific decision and validate that logic with business stakeholders.

• Reproducibility: We ensure that if we build a model again using the same artifacts and frameworks, we get the same result. The reproducibility of the model and its artifacts is critical for governance.

• Continuous monitoring: Once a model is in production, we constantly monitor for ‘model drift’. Even if a model works perfectly at launch, changes in demographics or data patterns can cause its performance to degrade over time. We track this rigorously and retune or retrain the models as required.

This framework ensures that we keep our clients’ data secure and manage our fiduciary duty effectively while leveraging advanced technologies.

Successful AI transformation requires a shift in culture as much as technology. How are you building internal capabilities and driving the necessary mindset change across the organisation?
You are absolutely right; challenges often surface when people treat AI purely as a technology problem. We have to start from the business lens. My role as Chief AI Officer involves three distinct functions: evaluate, enable, and educate.

Adoption is really about education. We approach this in three ways:

1. Workshops with external specialists: We bring in experts to conduct sessions on specific tools, ensuring teams understand the broader capabilities and limitations, as we did during our AIDLC rollout.

2. Knowledge sharing sessions: We host internal showcases to discuss new AI trends, the ‘personality’ of different tools, and when to use which tool. Understanding that each AI model behaves differently is key to effective usage.

3. Internal champions: We have identified and cultivated champions within the organisation who understand these technologies deeply. They act as bridges, educating their peers and driving adoption from within.

Looking ahead, what emerging AI technologies or trends do you believe will shape Motilal Oswal’s strategy over the next 3–5 years?
To look forward, we often have to look back. The industry journey started with data, moved to information, then to insights, and currently, we are at the stage of conversational insights.

The next frontier is ‘action models’. We are heading towards a future where a client can simply tell the system, ‘Please buy 500 Motilal Oswal shares at this price,’ and the AI will execute the entire transaction end-to-end. We are moving from a paradigm of searching for information to having the AI take action on our behalf.

Secondly, we will see a reduction in ‘search’ behaviour. Demographic trends already show that users are searching less and having conversations more. This conversational interface will become the dominant mode of interaction.

Finally, the evolution of compute power, including the potential advent of Quantum Computing, will play a significant role. In the coming years, we will have to consider how these advanced compute capabilities can further enable our long-term vision of simplicity, speed, and scale. We are at a juncture where AI will provide multi-fold advantages, doing things five times faster and driving three times the outcomes, and embracing this shift is the only way to create lasting competitive differentiation.