With the emergence of new technologies like Meta, there is a vast area in AI and Ethics that needs to be diagnosed in which human emancipation and empowerment play a crucial role, says Lakshmi Shastry, Principal Architect, Technology Advisory & Consulting, Brillio, in an interaction with CIO AXIS.
CIO AXIS: Give us a brief overview on how Brillio leverages AI.
Lakshmi Shastry: Over many past decades, businesses have been building their capabilities for analytics by relying on quantitative methods to support decision-making. In the present world, Artificial Intelligence plays a crucial role by doing more than just compute data. Significant advances and sophistication helps to analyse, predict, and derive insights that are crucial for businesses. It is one of the superpowers that will drive the world and Brillio is focused on building capabilities for the future and integrating AI into business growth models by delivering cutting-edge digital solutions with intelligent workflows to transform user experiences, significant efficiency and productivity, increased accuracy for smarter decisions, personalized communications across enterprises and industries. With our recent acquisition of Standav and Cedrus Digital, we are growing our team of specialists ready to help our customers accelerate business transformation and navigate through the evolving AI landscape.
CIO AXIS: Elaborate on some of AI governance methods, techniques, or frameworks used within your organisation to ensure that your products/solutions provide the best possible experience to the users with real-life examples or use cases.
Lakshmi Shastry: We have embraced a human-centric approach with the adoption of ethical principles that facilitate innovation while safeguarding trust, to ensure the positive impact of AI. Brillio’s Responsible AI is an “optimized” platform-agnostic approach to provide fair & transparent AI solutions to Customers. Model performance and model behaviors are continuously evolving with internal governance measures, augmented decision making& operations management to deploy Responsible AI.
Data Governance is ensured across Data Security, Loss Prevention, Integrity, Lineage, and Completeness. Process Governance is achieved through formalizing steps in ML lifecycle, Human interventions, Validation Checks with Reviews, Supporting Materials and Model Governance through Versioning for traceability, experiment tracking for an appropriate model, continuous integration, and continuous deployment.
CIO AXIS: Why are tech leaders talking about AI ethics, responsibility, and fairness in the present era? Why is it the need of the hour? How can it be ensured at scale?
Lakshmi Shastry: Exponential growth in data and computing power has fuelled the advancement of data-driven technologies. A major concern with today’s technological intervention is the question of accountability. AI systems have tons of data and ethics in AI gained attention and urgency due to its rapid development during the past decade. Ethics in AI is a critical topic in the present times because experts are trying to introduce transparency, justice and fairness, privacy, and responsibility that are essential to drive the moral machine of Artificial Intelligence. With the emergence of new technologies like Meta, there is a vast area in AI and Ethics that needs to be diagnosed in which human emancipation and empowerment play a crucial role.
CIO AXIS: How are you ensuring that your data science and AI/ML teams at Brillio are aligned with the Brillio’s AI governance policies, or best practices? Give example.
Lakshmi Shastry: As with any new technology, AI also introduces risks of unintended discrimination potentially leading to unfair outcomes. Brillio strives to identify the risks, tailor measures to address these risks, assist to build stakeholder confidence, adopt accountability-based practices with good data management and protection for responsibly deploying AI. AI solutions are made more explainable, unbiased, and trustworthy. As adopting AI-based systems for decision-making will increase significantly, Brillio’s – Responsible AI comes into play for our customers, allowing us to reap maximum benefits out of it.
CIO AXIS: Research has identified nearly 200 biases that influence human decision making. How do you avoid those biases from being introduced into your AI algorithms?
Lakshmi Shastry: Our primary consideration during the design, development, and deployment of AI is to adhere to a few guiding principles to promote trust in AI. With the advent of highly sophisticated & accurate AI model-building techniques, we are cognizant of the fact that predictions made by these AI models should be unbiased and explainable, to enable businesses to make righteous high-stake data-driven business decisions. As AI is used to amplify human capabilities, we adopt protection of interests, well-being & safety as guidance aid.
CIO AXIS: Do you have a due diligence process in place to ensure that data is collected ethically, especially while using third party plug and play data sets or models?
Lakshmi Shastry: Our layered governance and Bias Assessment Framework acts as sliding doors to assess the current state and map gaps quickly. Important performance scores are captured across Strategy, Governance, Data Bias, Algo Bias, and Explainability. Strong governance is key to achieving the fairness and trustworthiness needed for Responsible AI. Our Data Governance, Process Governance, and Model Governance elevate maturity.
CIO AXIS: How do you systematically feed ethical principles related to AI and AI applications into your platform?
Lakshmi Shastry: Our model design framework is based on two key principles of Responsible AI, AI Fairness and Ethical AI, and supported by underlying Justness and Transparency, Governance, and Privacy. Brillio approaches this by tackling bias in data and algorithms through custom-built AI/ML pipelines and determines the level of human involvement required in AI-augmented meaningful decision making.
CIO AXIS: How does Brillio ensure the protection of consumer data privacy?
Lakshmi Shastry: Like a domino effect of digitization, organizations are now more vulnerable to cyber threats than ever before. Hackers are also becoming more intelligent, trying newer ways of data theft causing business disruption. To help our customers protect their assets against such risks, we have adopted a seven-layered security approach to be continuously aware and defend from any potential threats. Digital security, adoption of zero-trust architecture, safeguarding networks, securing endpoints, adopting DevSecOps, adhering to data protection policies, and identity management are the multiple aspects of a robust security framework to ensure consumer data privacy.
CIO AXIS: What are your efforts in helping brands foster a trusted, transparent relationship with consumers?
Lakshmi Shastry: At Brillio, our efforts are focused on helping clients walk away from legacy IT and adopt emerging technologies. Built on the foundation of digital capabilities and through our customer-centric approach, we strive to deliver excellence for our customers and build a long-standing relationship. However, with the decline in customer trust in the present world, we ensure creating ethical AI frameworks, controls, and measures to prevent reputation loss and build customer trust. We continuously educate our workforce across levels on risks, mitigation, and the significance of adherence to the set standards. With set boundaries, Brillio is tapping into the full potential of AI to develop products best suited for our customers.
CIO AXIS: Did you come across any biases, or ethical concerns/issues lately within your organisation/industry/product? If yes, how did you address them?
Lakshmi Shastry: As we are in a fast-paced growth trajectory, we are making conscious investments into a lot of technology-led platforms, be it in hiring or healthcare, that has embraced data and knowledge base to balance fair, indiscriminate, diversity and inclusion based on merit & rationale. Our focus is to bridge misconceptions, with trusted AI that provides transparency and understanding.