Assistant to Streamline the IC Memo
Pre-Acquisition / Buy-Side Diagnostics
$2 BI
Nulla vitae elit libero
$4.3 MI
Fusce dapibus tellus
73%
Ipsum libero
Goal
The active equities manager aimed to streamline its memo drafting process with an interactive investment knowledge co-pilot to assist analysts in extracting and synthesizing insights from diverse documents.
Approach
• Mapped the investment process to define requirements for UI, workflow, response time, and data management.
• Built a capability to extract knowledge from diverse documents, tables, and public sources like Google, leveraging LLMs such as GPT-4.
• Collaborated with investment teams to ensure the co-pilot created differentiated insights from multiple sources and added value beyond simple information retrieval.
• Demonstrated the accuracy of outputs and the steps to achieve them, ensuring explainability of the plan, intermediate steps, and references.
• Enhanced the solution’s reliability and performance by limiting the search area.
Interventions
Delivered a set of LLM-agnostic capabilities that improved the performance of off-the-shelf LLM tools, embedding investment process knowledge into planning and execution agents.
The deployed preliminary suite included the following capabilities:
• Conversational interface for question answering across multiple files and company clusters.
• Summarization and key point extraction from multi-modal data sources and documentation.
• Proprietary planner agent to support interrogation and questioning by analysts (e.g., compare and contrast).
• Tool integration for retrieving information from the web and accessing third-party tools (e.g., Wolfram Alpha).
Additionally:
• Up-skilled the internal data science team on technical best practices through day-to-day collaboration and on-the-job coaching.