Relationship Managers (RMs) were spending up to 20 minutes manually gathering client portfolio and market data from internal dashboards and external sources. This fragmented process delayed meeting preparation and limited the quality of client conversations. The absence of a unified view across family and client-level data further increased reporting effort and reduced operational efficiency.

Exponentia.ai implemented a GenAI-powered chatbot solution that integrated structured data from AWS S3 and unstructured data from SharePoint. The chatbot enabled RMs to:

• Retrieve client portfolio insights using natural language queries.
• Access data seamlessly via Microsoft Teams and mobile devices.
• View consolidated summaries across families and individual clients.
• Eliminate manual inference and reporting efforts.

The solution leveraged Retrieval-Augmented Generation (RAG) and Text-to-SQL (T2SQL) web search for real-time stock market data & news, and multi-agent systems technologies to deliver precise, contextual responses in seconds.

‍Security and Compliance

‍The solution was deployed in a secure, VPC-enabled AWS environment with Databricks as the governed data platform.

Authentication was managed via Microsoft Entra ID and the client’s Prism app using short-lived OAuth tokens.

Unity Catalog enforced fine-grained access controls, ensuring RMs could only view data they were authorized to access.

All data was encrypted at rest and in transit, and every query was logged for auditability—enabling secure GenAI adoption at scale.

Quantitative

• 70–80% reduction in time spent retrieving client portfolio information.
• Time per client reduced from 20 minutes to ~5 minutes, with insights retrieved in 5–6 queries,     each taking 40–50 seconds.

Qualitative

• Enhanced client conversations with instant access to performance metrics, asset allocation, and market     trends.
• Improved meeting readiness and reduced manual reporting effort.
• Mobile and Teams-based access enabled informed, on-the-fly interactions.

Technology Stack

• GenAI Accelerator: AIXponent
• Data Sources: AWS S3 (structured), SharePoint (unstructured)
• Platforms: Microsoft Teams, Web Application
• Core Technologies: Retrieval-Augmented Generation (RAG), Text-to-SQL (T2SQL), Natural Language Processing (NLP)

‍