Lightkeeper has launched Lightkeeper Beacon, a new AI-enabled capability designed to allow hedge funds and asset managers to query their portfolio data in plain English while maintaining full transparency, governance and auditability.
Beacon is now available to all Lightkeeper clients and enables users to ask natural-language questions about portfolio performance, risk and analytics, receiving answers generated from Lightkeeper’s validated, institutional-grade data rather than unverified external sources.
The launch comes as asset managers increasingly experiment with large language models (LLMs) to speed up analysis, but struggle to connect those tools to clean, proprietary portfolio data in a controlled way. Beacon addresses this gap by operating directly on data that Lightkeeper already aggregates, validates and calculates for clients, ensuring that every AI-generated response can be traced back to a trusted system of record.
Beacon integrates client data with leading LLMs such as OpenAI’s ChatGPT and Anthropic’s Claude via the Model Context Protocol (MCP), an open standard that allows governed access to institutional datasets. This architecture enables firms to combine proprietary portfolio analytics with broader contextual reasoning from LLMs, without compromising permissions, controls or audit trails.
Lightkeeper said it deliberately chose to integrate its platform with external LLMs rather than embedding fixed AI features, giving clients flexibility to extend their own AI use cases while keeping all calculations grounded in established definitions and workflows.
The tool was developed in close collaboration with clients through roundtables and beta testing, with a strong focus on verification, transparency and operational fit. During the beta phase, one investment firm used Beacon to automate year-end analyst performance reviews, a process that typically took hours. By asking Beacon to perform a year-to-date performance analysis with qualitative commentary, the firm generated a detailed report that was incorporated directly into compensation discussions, significantly reducing preparation time.
Beacon is the first step in Lightkeeper’s broader AI roadmap, with the firm also developing Lightkeeper Lumina, currently in beta, which is intended to embed AI-driven intelligence directly into the platform to proactively surface insights.
