Automated code insights have improved security checks and made review workflows more consistentIf I talk about improvements for CAST Highlight, I would suggest three things. The first is better understanding or niche understanding. Right now, the intent matching is strong for general topics such as ease of use, but for niche B2B terms such as HIPAA compliance or multi-tenant architecture, it sometimes misses context. Improving the domain-specific models would make highlights more accurate for these verticals. The second improvement is more control over the deduplication logic. CAST Highlight's deduplication is great for avoiding spam, but sometimes we want two similar quotes if they are from very different company sizes, such as SMB versus enterprise perspectives on pricing. A slider to adjust deduplication strictness would help. The third suggestion I would like to give is deeper sentiment and outcome tagging. While it has core sentiment capabilities, it does not tag outcomes automatically. For instance, if a quote mentions saved $50,000 per year, tagging that as cost savings $50,000 would let us build ROI charts instantly instead of reading each quote manually. Regarding user experience, integrations, and reporting, I think there is room to enhance those aspects. Regarding user experience, I would suggest improving user actions in terms of bulk actions and keyboard shortcuts. Day-to-day analysts review 50-plus suggested quotes, and currently it is mostly clicking to approve one by one. Adding bulk approve or reject options and keyboard shortcuts would significantly reduce the time taken. A small UX change can lead to a big speed boost. The second point is integrations when pushing to the CMS and Slack alerts. Right now, we export approved highlights manually from CAST Highlight. If CAST Highlight could push directly to our CMS or send Slack alerts for high-strength quotes that hit trending topics, it would close the loop faster, reducing copy-pasting. The third improvement relates to reporting, specifically custom insight dashboards. The tool displays which topics have the most highlights, but we cannot build custom dashboards yet. For example, showing all security quotes from healthcare companies with more than 1,000 employees over the last 90 days would enable better filtering, and exportable dashboards would streamline quarterly reviews.
