Sciene AI Companion: building an autonomous Customer Success platform on Databricks

Summary

Sciene built AI Companion for Quartile’s Customer Success organization, where CSMs support more than 1,000 brands and previously spent substantial time preparing decks, reconstructing context, and investigating account changes. The platform addresses personalization at scale, high-volume content generation, and root-cause diagnosis by combining account data, CSM communication styles, company principles, and cross-domain business data. Its Email Hub cuts reply time from 15–30 minutes to about three minutes, Meeting Hub reduces preparation for 80+ slide decks from over two hours to around 10 minutes, and Account Flagging reduces diagnosis of flagged accounts from 30+ minutes to about five. Databricks provides the shared governed foundation: Delta Sharing supplies data without copies, Lakebase stores operational state, and SQL Warehouses serve analytical, AI, and operational workloads from the same tables. The design keeps CSMs responsible for judgment while giving them current context for customer interactions.

Context

Quartile’s Customer Success Managers needed to support more than 1,000 brands across geographies and time zones, but assembling presentations, reconstructing account context, drafting replies, and investigating changes consumed substantial time. The platform also had to preserve personalization, handle high content volume, and diagnose root causes across advertising, campaign, inventory, billing, and CRM data.

Approach / What changed

Sciene built AI Companion with Email Hub, Meeting Hub, and Account Flagging modules on Databricks. Delta Sharing provides governed data without copies, Lakebase stores operational state and generated content, and SQL Warehouses support analytical, AI inference, and operational workloads from shared tables. The modules use current account data, communication styles, company principles, meeting history, and cross-domain evaluations to produce drafts, decks, and ranked diagnostic alerts.

Takeaways

• Email Hub reduced surveyed reply time from 15–30 minutes to about three minutes, while Meeting Hub reduced preparation for an 80+ slide deck from over two hours to around 10 minutes.
• Account Flagging evaluates advertising, campaign, inventory, billing, and CRM data daily, writes severity-ranked alerts to Lakebase, and supports threshold changes through configuration rather than code releases.
• Using shared governed tables prevents separate analytical, AI, and application data stores from developing synchronization drift; CSMs retain ownership of account decisions and customer relationships.

Related reading