As Indian organizations seek to enhance efficiency and drive business outcomes, the adoption of artificial intelligence (AI) remains a key priority. However, aligning AI strategies with broader business goals and overcoming challenges such as data silos, quality issues, and organizational readiness is crucial for success.
In this exclusive interview, Rohit Ramanand, Group Vice President of Engineering, India at New Relic, shares his expert insights on how observability plays a pivotal role in AI readiness, the obstacles Indian companies face on their AI adoption journey, and practical steps for overcoming these hurdles. Rohit also discusses how New Relic is working with Indian businesses to bridge the gap between data quality and AI maturity, providing actionable recommendations for organizations looking to future-proof their AI strategies.
Despite the potential of AI, why do you think less than 15% of organisations in India have aligned their AI strategy with broader corporate goals? How is New Relic helping businesses bridge this gap and align their AI strategy with their broader corporate objectives?
There’s no denying how challenging AI implementation can be. A recent study by Gartner found that nearly a third of all AI projects will be abandoned at the proof of concept stage. Most often, AI implementation fails because businesses don’t align their AI strategies with clearly defined goals. In many cases, the broader corporate goals themselves aren’t clear enough for companies to effectively measure the outcome of AI implementation. The first step is understanding the problem sets and implementing AI in the areas that drive better value. Merely implementing AI to keep up with hype cycles doesn’t translate into positive ROI.
Once AI is strategically implemented to solve critical problems within a workflow, New Relic steps in to help businesses observe AI systems. New Relic’s intelligent observability platform monitors the performance of AI systems, detects errors and aids in troubleshooting. It monitors for bias, toxicity, and hallucinations in complex AI models to ensure fair and reliable outcomes, and can also identify and help resolve computational bottlenecks, ensuring AI apps are responsive and efficient.
Furthermore, in a cost-conscious market like India, New Relic ensures businesses optimise costs. Our New Relic AI Monitoring solution tracks token processing to manage AI model costs effectively and stay within budgetary limits. Additionally, it monitors AI applications for vulnerabilities to mitigate potential security attacks. While we help businesses optimise AI spend and smoothen the implementation process, we also help measure whether an AI strategy is actually working, offering a feedback loop that businesses can feed back into the system that can then constantly improve.
Can you elaborate on the common roadblocks and challenges Indian companies face when trying to adopt AI technologies? How does New Relic work with these organisations to overcome these hurdles and accelerate their AI adoption journey?
Data is fueling AI projects. Without robust data, AI models operate as garbage in, garbage out productions, resulting in AI projects that struggle to produce accurate and reliable results. According to a NewVantage survey, 63% of organisations say that data quality is a big challenge. Moreover, a Vanson Bourne survey revealed 73% of organisations find translating data insights into practical advice for decision-makers a challenge. The numbers make it clear that there’s an urgent need for effective data management and monitoring. Additionally, businesses have to contend with model drift, where AI models become less effective as the data distribution changes over time.
Many AI models, particularly deep learning models, are black boxes. They make it difficult to understand the AI’s decisions. New Relic solves many of these problems by helping companies accurately measure the performance of their AI projects. AI monitoring can actually help manage and monitor AI models and really tell businesses whether they’re functioning optimally or not and whether they’re able to actually scale the AI apps to the levels they want.
Observability goes beyond traditional monitoring, but how does it specifically contribute to making businesses AI-ready? What role does New Relic’s observability solutions play in helping Indian businesses gain the insights needed for successful AI implementation?
To be AI-ready, businesses need systems in place to detect issues, rather than focus on them reactively. The business must be able to use data for troubleshooting and debugging, optimise the performance of systems, and analyse the data to understand what value they can derive from it.
For example, if a company wants to use data to personalise customer experience, they would also need visibility and the analysis of the data itself, which can offer insights into whether the business will actually be able to achieve their goals with existing models or not. These feedback loops are crucial to making AI accurate and eliminating bias, and as AI improves over time, no model will be 100% accurate from day one. This feedback loop relies on the quality of data provided and the amount of feedback it’s able to generate to know if something is going wrong, and correct it.
New Relic plays a big role in telling businesses when something’s not right, and that they need to go back and provide feedback. Additionally, it is instrumental in driving better user experience. Many AI systems today are starting to interact with end users like chatbots. Our platform helps companies monitor and analyse these interactions, helping businesses understand pitfalls, so they can proactively improve UX.
What practical steps can Indian companies take to eliminate data silos and improve data quality for AI operations? How is New Relic supporting organisations in their efforts to break down these silos and ensure high-quality data flows across their systems?
Businesses must begin by putting in place a data governance strategy. It requires answers to some critical questions, including: Where does the data flow from? Where does the data go? Who has access to the data? What is the source of the data, and does it have multiple sources? A unified data repository or a single source of truth brings all data together in a single place, as it becomes easier to access, aggregate, store, manage, and analyse. Once organisations have this strategy in place, they need the right tools to be able to access the data and transform it for better business outcomes.
Data standardisation is central to effective AI implementation. Without it, maintenance would become a nightmare when different tools are operating on different data sets. That’s where New Relic helps. Our platform’s ability to ingest data from any kind of system, along with open telemetry ensures that businesses can plug it into their systems and it can pull all of the data and bring it into one centralised repository. New Relic’s solutions are great at data integration and visualisation, helping eliminate silos. Additionally, it automates the data management process, offering insights into what’s going wrong with systems, where it’s going wrong, and how to fix it. It also ensures that data is compliant, so businesses don’t get into trouble with regulators.
What are the best practices you recommend for organisations looking to strategically align AI with their business goals?
It’s really about having clear business objectives. Without it, businesses wouldn’t know what problem AI is meant to solve. Blindly implementing AI without understanding whether it would improve business outcomes results in poor ROI. The best practice is to identify key goals and align AI initiatives to those goals, while also having mechanisms to measure the success of the AI.
Furthermore, AI implementation isn’t possible without leadership buy-in. Executive sponsorship and leadership ensure that goals are aligned, so it effectively trickles down across the organisations. While it is important to have a vision, it is also important to have a plan of action. It’s also essential to build the right talent. Without the right skillset, it’s not possible to scale AI initiatives. The onus is on businesses to upskill their people and have a clear upskilling strategy to ensure teams have the right people to implement AI.
AI has innumerable use cases. The biggest pitfall is prioritising use cases. It’s very important to have a strategy that determines which use cases will work for the organisation as AI isn’t a one size fits all proposition. Because adopting the wrong tool can hamper the ability to scale. Additionally, as a part of the strategy, there must be mechanisms to define success metrics and monitor them, as it helps quickly pivot goals, reset and redefine them to make the most of the AI tools and also build trust in these systems.
What trends do you anticipate, and how can Indian businesses prepare for the future of AI?
It’s no surprise that the adoption of generative AI will expand, and it’s going to be used for everything from creating content, to automating tasks, and generating new ideas. It’s important to invest and prepare for the skills that will be required to explore more use cases of generative AI. This brings us to AI governance. Ethical governance is going to become a critical trend going forward. Companies that will invest in technology that can evaluate AI models to determine whether they are ethical or not.
We foresee the proliferation of technologies that are focused on assessing the efficacy and accuracy of AI models. Today, a lot of the AI-led initiatives are focused on personalisation at scale. Currently, most of the personalisation processes are occurring offline. With greater AI adoption, businesses will invest in solutions that offer edge personalisation. For example, there could be a day when businesses offer personalised pricing based on a customer’s shopping history.
Being prepared for these changes requires businesses to invest in building skill sets either by hiring new talent or upskilling existing talent. Businesses must ensure they have the right strategy in place, so they can translate it into measurable, achievable goals, to ensure they are building use cases for AI that are going to be valuable to their customers.
Businesses must prepare by understanding how AI is going to help them make money, and ensuring there are regular audits that you keep conducting to ensure they stay relevant. It’s essential to install feedback loops so the AI models keep getting better as the landscape is changing very quickly. It’s also important to invest in the right technology and infrastructure necessary to scale AI initiatives. Businesses must begin preparing these basic building blocks to unlock true value.
