Title

AWS re:Invent 2022 - How Moody’s uses serverless and microservices for ESG scores (FSI205)

Summary

• Introduction: Sri Santaneni and Divya Lati from Moody's ESG discuss how they use serverless and microservices to generate ESG data for financial services.
• Moody's Overview: Moody's is a global risk assessment firm providing credit ratings and financial risk assessment solutions. Moody's Analytics offers analytics solutions for risk management and growth.
• ESG Overview: ESG stands for Environmental, Social, and Governance. It's not just a regulation but a reputational risk for companies, influencing investment funds and bank lending practices.
• Challenges: The challenges in ESG data collection include lack of standard reporting, infrequent reporting, and difficulty in uncovering insights from the data.
• Financial Incentives: Sustainable funds are performing better, and there is a business case for providing ESG data.
• Moody's ESG Data Journey: Divya Lati explains the five-step process Moody's uses to gather ESG data, which includes laying the foundation for ESG reference data, collecting publicly disclosed documents, extracting information using AI/ML, enriching and vetting the data, and finally scoring and distributing the data.
• Architecture: The architecture is serverless, leveraging AWS services like Glue, Lambda, DynamoDB, and step functions to handle various aspects of the ESG data journey.
• Resiliency: The system is designed for resiliency with goals for recovery time objectives and data durability.
• Outcome: The product was launched in Q1 2022, processing tens of thousands of entities and hundreds of thousands of documents, with each entity having over 300 metrics scanned for a final ESG score.

Insights

• Serverless Benefits: The use of serverless architecture and microservices has provided scalability, reduced code duplication, and ensured data lineage and consistency across Moody's systems.
• Data Lineage Importance: Emphasis on data lineage throughout the process ensures transparency and traceability of ESG data, which is critical for credibility in the financial services industry.
• Challenges in Standardization: The lack of standardization in ESG reporting presents a significant challenge, requiring sophisticated AI/ML and manual processes to extract and validate data.
• Continuous Improvement: The AI/ML models require active learning and continuous feedback from analysts to improve accuracy and efficiency in data extraction.
• Business Case for ESG: There is a clear financial incentive for companies to invest in ESG data collection and reporting, as sustainable funds show better performance and create new revenue streams.
• IT's Role in ESG: The IT sector can contribute to ESG efforts by creating sustainable code that minimizes data center resource consumption.
• Resiliency as a Key Factor: Building a resilient system is crucial, especially in handling the complexities of data pipelines and ensuring uninterrupted service to global users.
• Product Impact: The successful launch of the product demonstrates the practical application of AWS serverless technologies in addressing complex data challenges in the financial sector.