Worked on the InferESG repository to deliver prompt improvements for ESG report generation, focusing on enhancing report clarity and contextual relevance. Refactored report headings to use a unified report_heading structure and updated prompt templates to ensure consistency across outputs. Integrated industry sector data collection, enabling the generation of richer markdown outputs that provide stakeholders with more actionable business context. The work emphasized backend development and code refactoring, leveraging Python and Jinja2 to streamline template-driven prompts. No major bugs were addressed during this period, as the primary focus remained on feature delivery and maintainability to support better team alignment and decision-making.