Aurionpro unveils AI explainability platform

Tech solutions company Aurionpro’s unit, Arya.ai, has introduced AryaXAI, an artificial intelligence (AI) explainability platform seeking to improve transparency and accountability in critical sectors, including finance, healthcare, and defense.

At the core of AryaXAI is DLBacktrace, a newly developed technique for deep learning models that enables enterprises to trace the decision-making processes of AI systems. Arya.ai has made DLBacktrace open-source, allowing the global developer community to access it through GitHub.

The platform includes a set of tools for explainable AI (XAI), integrating established methods like SHAP, LIME, and Integrated Gradients. These tools help users choose the most effective approach for explaining the behaviour of their specific AI models.

Vinay Kumar Sankarapu, CEO of Arya.ai, emphasised the platform’s importance in addressing the increasing complexity of AI systems. He highlighted the need for clear insights into AI decision-making to foster trust and meet evolving regulatory requirements.

AryaXAI also includes monitoring capabilities to detect issues such as data drift and model bias, along with customisable risk management frameworks. It offers synthetic alignment strategies and automated logging features to support the deployment of robust and responsible AI systems.

In April, Aurionpro Solutions acquired majority stake in Arya.ai, a Mumbai-based startup specialising in banking and insurance platforms, for approximately $16.5 million.

This development underscores the growing emphasis on explainable and accountable AI in industries where decision-making processes must be transparent and reliable.