The National Insurance Information System (NSIS) is developing a comprehensive strategy to integrate artificial intelligence into its core business operations. The company plans to utilize classical machine learning, generative neural networks, and computer vision to optimize actuarial calculations, data quality, and fraud detection.
Currently, NSIS employs AI models to process natural language requests and extract data from scanned documents. Alexander Makarov, head of the AI development department, stated that the company intends to implement document verification tools to detect forgeries, alongside existing recognition systems for passports and vehicle documents.
The strategy includes the development of ML models for automated data cleaning, duplicate merging, and the detection of fraudulent schemes using graph analysis. Additionally, the company is focusing on 'explainable AI' to ensure transparency in decision-making processes and maintain strict information security standards.
Lead system architect Dmitry Bolot announced plans to introduce an AI assistant for developers to accelerate coding, code reviews, and documentation generation. The company will also implement automated architectural inventory systems to validate code changes and identify potential security and quality issues early in the development cycle.
