AI Initiatives cannot be complete if it does not comply with a regulatory environment: Puneet Asthana, ED & CTO, Shriram Wealth
In an exclusive interview with Express Computer, Puneet Asthana, Chief Technology Officer of Shriram Wealth, explains how the organisation is gearing up to revolutionise the wealth management future by adopting AI. He presents an end-to-end strategy involving having a strong layer of data consolidation, compliance in accordance with RBI, SEBI, and DPDP regulations, and investing in scalable, secured hybrid-cloud environments. He also highlights the importance of ethical and explainable AI, supported by customer consent engines and transparent data practices. Asthana also speaks about developing real-time AI models for risk assessment, dynamic portfolio optimisation, and effective wealth management services. Looking ahead, he identifies federated learning and blockchain as transformative technologies that will drive industry-wide trust and collaboration. With a strong focus on data accuracy, customer-centric design, and adaptive learning, Shriram Wealth is positioning itself as a leader in responsible AI adoption across the financial services sector.
With AI playing an increasingly pivotal role in financial services, how is Shriram Wealth leveraging data and intelligence to deliver truly personalised client experiences at scale?
I will break it in three steps. Firstly, we focus on generating deep customer insights, which stem from creating a unified data layer that consolidates disparate datasets across our organisation. This consolidated data becomes the backbone for advanced analytics, enabling us to micro-segment our customer base with precision. Second, we construct AI agents that reside on top of this data foundation so that we can apply predictive analytics and produce actionable insights specific to the profile of a client. Finally, we make our entire tech environment real-time and scalable. This is crucial in the current environment where one needs to make decisions immediately, not after batch-processing it overnight. By designing our systems for responsiveness, security, and compliance, we are able to provide trust and personalised value at scale.
Beyond operational efficiency, how is AI shaping strategic decision-making in wealth management — especially for dynamic portfolio optimisation and long-term services?
AI has evolved from being a backend efficiency driver to a front-line strategic enabler. In terms of portfolio optimisation, we use AI models that ingest real-time market data alongside predefined business risk matrices. These models continuously learn and adapt, offering what we call adaptive learning. For instance, they dynamically rebalance portfolios and provide recommendations in real time, usually before the market closes. On the services front, we create scenario-based models that model future market scenarios like a shift in CRR or a merger in the steel industry and simulate portfolio reactions to these hypothetical scenarios. These simulations enable us to offer data-driven wealth management services that are more than gut-feel or sentiment of the market. AI assists us in parameterising risk, forecasting future states, and recommending bespoke investment strategies based on changing client objectives and market conditions.
How would a future-proof AI platform for a wealth management company look from the point of view of a CTO? How do you make it scalable and agile in the face of rapid technology changes?
I feel that a future-proof AI infrastructure needs to be developed on three core principles; scalability, modularity, and flexibility. Today’s world runs on cloud, microservices, and containerisation, and we are no exception. I advocate for an API-first architecture, because soon models will not only interface with applications but with each other like think of a risk model interacting directly with a customer behaviour model. This is where federated learning will play a major role. Also, considering regulatory constraints in BFSI, our infrastructure must work seamlessly across hybrid cloud environments. Certain workloads must remain on-premise due to compliance, so cross-cloud portability becomes essential. Above all, every layer from data handling to AI model deployment must align with SEBI and RBI’s cybersecurity mandates. Innovation is necessary, but not at the expense of compliance or ethical integrity.
Given the sensitive nature of financial data, how do you ensure that AI solutions at Shriram Wealth remain transparent, ethical, and compliant — without stifling innovation?
We strongly integrate explainability and ethics principles into our infrastructure. If an AI model recommends a portfolio rebalance, it also needs to explain the ‘why’ in a language understandable by humans — that’s what explainable AI is all about. We are also heavily investing in developing a customer consent engine that complies with DPDP regulations. Customers need to have complete visibility and control over what data they’re providing and how it is utilised. We are also developing a federated consent dashboard, enabling customers to monitor and manage their consents between financial institutions. Ethics, transparency, and compliance are not mere trade-offs, rather they are enablers of trust. We think customers will be more confident in adopting AI if they have a sense that it’s working for them and with full transparency.
Looking ahead, which AI-driven trends or emerging technologies will redefine wealth management in the next 3–5 years and how is Shriram Wealth preparing for this evolution?
The next few years will see AI models becoming more collaborative and decentralised, particularly through federated learning. For instance, fraud detection today happens in organisational silos. Federated models can train on data from multiple brokers or banks while maintaining privacy, allowing the industry to detect anomalies system-wide. Also, the role of the Data Protection Officer (DPO) is becoming pivotal — every firm will need one to oversee ethical use and data compliance. Beyond AI, I foresee blockchain re-entering the spotlight but not as a hype term, but as a backbone for inter-organisational trust and data lineage. When federated ecosystems are the standard, distributed ledger technologies will be necessary to impose transparency and immutability on players.
How is Shriram Wealth using AI to make everyday work easier for clients and employees?
Our current focus is on empowering our relationship managers (RMs), who remain the cornerstone of our wealth business. Through building a unified data lake that integrates customer information, market trends, and internal risk factors, we equip RMs with tools that provide real-time insights and smart recommendations. Imagine an RM sitting with a client and querying a model that instantly suggests investment strategies tailored to that client’s risk appetite, life goals, and current portfolio; that’s what we’re enabling. Internally, we’re also working on customer segmentation and smart analytics, allowing employees to understand client behaviours deeply and respond with precision. It reduces guesswork and enhances productivity, accuracy, and customer satisfaction.
What role will AI play in helping Shriram Wealth differentiate itself in the wealth management space?
AI will be our key differentiator. We’re not retrofitting AI onto old systems; we’ve designed our wealth business with AI at its core from day one. Speed, accuracy, and trust are our three pillars. Our data strategy which includes high-integrity data handling, intelligent model development, and transparent communication that enables us to offer precise, trustworthy, and tailored solutions across customer segments. We’re not just serving HNIs and UHNWIs; we’re building AI tools that make wealth advice accessible and relevant to all client tiers. Our USP is democratising intelligent financial services powered by ethical, robust AI.
