FAQ

CapRelay helps financial professionals research U.S. public companies, generate ideas, and stay current on companies of interest. CapRelay offers written research, like an investment bank's research department, with more modern capabilities.

CapRelay is based on public SEC filings, earnings call transcripts, and other proprietary data. We clean, process, and analyze this data with cutting-edge AI models, with a focus on accuracy and relevance. As investors ourselves, we understand the importance of reliable information. Each company research page includes details on the underlying sources.

We teach AI systems how to think like an analyst. Our system replicates how an analyst thinks, in discrete steps. When given a new company, analysts wouldn't jump into figuring out the bull-bear debate or bear case, but rather, start by learning what the company does, what happened historically, studying the trends and news items that impacted the company over time, etc. Then, using these as foundations, the analyst can drive towards interpreting the bull case, bear case, and ultimately the debate. Our system replicates this analyst process by teaching different agents to execute different analyses with different context: one agent analyzes risks, another analyzes the history, and so on, before combining it in downstream, more difficult analyses.

We continuously ingest new documents in real-time and generate updated research as new source documents become available. Most content (including earnings reviews) is generated several hours after earnings calls end. Other content (such as management and incentives analyses) are based on the proxy, and are therefore generated annually after the proxy is filed.

Screener Search uses natural-language queries and AI-powered search to find companies based on CapRelay's proprietary research library. You can combine Search with Filters to narrow results by market cap, industry, coverage group, or thesis components. After running a Search, AI Review (experimental) can provide a deeper review of candidate companies, label each one as a match or not a match, and explain why - all grounded in CapRelay's research library.

Yes. All CapRelay subscribers have MCP access, which lets AI agents like Claude, Claude Code, Codex, ChatGPT, and other OAuth-capable MCP clients to search for companies, use the Screener, and retrieve CapRelay research from chat. See the MCP installation guide for setup steps. We have also designed the site to be navigable for agents and include agent-friendly features like Copy for LLM, which copies clean markdown from supported research pages.

For higher MCP usage limits or enterprise-level access, please contact us.

No. Your MCP client initiates authenticated requests to CapRelay and receives CapRelay data in response. The connection does not give CapRelay credentials or direct access to your organization's applications, systems, or data stores, so CapRelay cannot access, create, modify, or delete data there. Any permissions granted to other tools connected to your agent are managed separately.

CapRelay applies usage limits to prevent bulk export of our intellectual property, prevent scraping or automated access, and protect the Service for customers. These limits are scoped to ordinary, human-scale usage, and human users are very unlikely to encounter them during normal use. If you believe you were limited incorrectly or need higher-volume programmatic access, please contact us.

We cover publicly traded companies in the U.S. Many foreign companies with U.S. ADRs or listed securities may also have coverage on CapRelay, though availability of data may limit the coverage provided. Some companies may not be available if their data is incomplete.

No. CapRelay provides research and analysis for informational purposes only. We do not provide investment advice. See our Terms of Service for more details.

Please reach out to us at - we'll get back to you as soon as possible.