New capability empowers regulated industries with explainable AI to enhance compliance, reduce risk, and optimize operational efficiency.
Menlo Park, CA — December 3, 2024 — Skan AI, a leader in AI-driven business process intelligence, today announced an advanced hybrid predictive capability, InferNet, within its Process Intelligence platform. InferNet represents a significant advancement in explainable AI, particularly benefiting regulated industries such as banking and insurance. This breakthrough demonstrates Skan AI’s continued commitment to advancing proprietary AI technology and represents a major leap in explainable AI. The announcement comes on the heels of the acceptance of the company’s research at the 5th International Workshop on AI-enabled Process Automation (AI-PA). Skan AI is a proud leader of cutting-edge advancements in AI for process automation.
Organizations today face increasing pressure to optimize workflows, reduce bottlenecks, and ensure compliance with stringent regulations. InferNet introduces a transformative approach by combining transformer-based models with Dynamic Bayesian Networks (DBNs), enabling enterprises to achieve high predictive accuracy while maintaining interpretability. This proprietary capability addresses critical challenges faced by regulated industries—particularly banking and insurance—by offering transparency that reduces regulatory risk and enhances operational efficiency. Recent industry challenges underscore the need for explainable AI solutions. For instance, financial institutions have faced large regulatory fines and reputational loss due to opaque AI-driven credit models that inadvertently reinforced biases, while some insurers have faced legal issues stemming from non-transparent AI algorithms used in claim approvals.
With a proprietary approach, InferNet, Skan AI addresses these challenges by pairing the predictive power of transformers with the interpretability of DBNs. This equips businesses with deeper, more accurate insights into complex processes while ensuring they understand the “why” behind each prediction. By capturing dependencies within workflows, DBNs provide transparency needed for informed decision-making, while transformers provide scalability and precision. Skan AI’s score-based marginalization further enhances InferNet’s adaptability by empowering it to handle new and unseen data combinations effectively—a common challenge in dynamic enterprise environments.
"Our goal with InferNet is to redefine what predictive process monitoring can achieve for regulated industries," said Avinash Misra, cofounder and CEO of Skan AI. "In banking and insurance, where compliance and customer trust are paramount, it’s important to understand not just the 'what' but the 'why' behind AI-driven recommendations. InferNet delivers a solution that drives both operational efficiency and transparency, empowering businesses to optimize with confidence."
The company’s commitment to advancing the boundaries of AI is evident in its use of transformers alongside DBNs. Manish Garg, cofounder and Chief Product Officer, noted, "Our commitment to advancing AI R&D is at the core of our mission to transform process intelligence. InferNet reflects our leadership in explainable AI, setting a new standard for what organizations should expect from predictive insights. With this capability, we’re ensuring that our customers not only benefit from powerful analytics but also from an AI they can trust and understand."
Skan AI Accepted to ICSOC 2024
Skan AI’s dedication to innovation is further recognized internationally: the company’s paper on the next likely action prediction has been accepted for presentation at the 5th International Workshop on AI-enabled Process Automation (AI-PA), held in conjunction with the 22nd International Conference on Service-Oriented Computing (ICSOC 2024). “Being recognized at AI-PA and ICSOC 2024 is a testament to the innovative strides Skan AI is making in AI-driven process intelligence,” added Dr. JC Bose, Chief Data Scientist at Skan AI. “With InferNet, we are advancing predictive AI by blending powerful frameworks like DBNs and transformer-based models, leading the way in creating transparent, reliable tools that enterprises can trust. I am really proud of the innovation from Gautam Vashishtha and the AI team." Dr. JC Bose is also presenting his research on December 3rd at the ISCOC 2024 conference.
Skan AI’s recent introduction of real-time execution visibility for Anti-Money Laundering (AML) programs aligns with the launch of InferNet, supporting financial institutions in compliance and fraud prevention. By providing transparent, real-time insights, these solutions empower organizations to detect irregularities, mitigate risks, and avoid costly fines or reputational damage.
Skan AI continues to lead the way in explainable AI, delivering powerful, proprietary innovations that meet the evolving demands of regulated industries while advancing the global conversation on trustworthy AI.