- Runtime Path, Founder and CIO
- Aravt Global, Partner
- Goldman Sachs, Analyst, Special Situations Group - Multi-Strategy Investing
Scale Up
Your Investment Research
Institutional-grade proprietary data and tools for investors and AI agents
How Investors Use CapRelay
Generate Differentiated New Ideas
- Discover new ideas by pattern-matching to existing theses, or screen based on your qualitative investment criteria.
- Determine whether an idea fits your framework in minutes, with buy-side grade bull and bear cases and thesis highlights.
- Investors can now learn enough to be conversant in minutes, enabling new strategies for idea gen.
Cover More Companies
- Earnings and event recaps comprehensive enough to skip the transcript, yet skimmable enough to read from your phone.
- Shortcut common research tasks like proxy review, call prep, industry read-across, and more.
- Get alerts for news events and summaries delivered to your inbox.
How Agents Use CapRelay
Power Advanced Automations
Give your AI agents the context and tools they need via MCP:
- Raw source data, including filings, transcripts, and podcast data, optimized for AI agents.
- Historical financial data.
- Buy-side quality research backed by custom post-trained models and differentiated data sources unlocks complex research prompts.
- Tools not available on the web, including screening, podcast analysis, search over buy-side research.
More Accurate Answers with Less Wasted Tokens
- Previously bespoke analyses can now be scaled across an entire investable universe with AI agents, making efficiency and reliability key.
- Don't let your agents spend time and tokens reinventing the wheel, like manually assembling financial history or recapping earnings. Get it in a one-shot request from CapRelay.
- Avoid error-prone agent tasks, like parsing SEC HTML tables or reading consumer finance websites.
Buy-Side
Insights and Data
CapRelay's custom post-trained models mirror how buy-side investors think.
- Filter out the noise to focus on what matters, across a vast corpus of information.
- Bull-bear debate example (left): Key tension points the investment community is focused on.
- Data is optimized for MCP delivery to underpin advanced idea generation or research automations.
Idea Generation Tools
Find ideas that match your investment criteria, with qualitative screening and pattern-matching tools, designed for both agents and humans.
- Surface under-the-radar opportunities by pattern-matching based on theses and themes.
- Example (image to the right): Which smaller company is running the playbook that made HEICO and TransDigm monster compounders?
- Search in natural language for companies that match your qualitative framework.
- Screen for companies that match your qualitative framework, e.g. consolidated market structures, secular growth, and more.
Podcast Coverage
Find relevant podcasts for company research, along with analysis of what's incremental to the buy-side bull and bear case.
- Find and analyze podcasts featuring company executives, industry experts, investors, and more.
- Get weekly digests for new podcasts on companies you watch.
- Pull data into Claude or ChatGPT via MCP to add more sources to your research.
Power AI Agent Workflows
With CapRelay's MCP
Analysts need the right context and tools to do great work. So do agents. CapRelay's MCP gives agents access to tools and institutional-grade data to unlock powerful automations. Examples of prompts CapRelay unlocks:
- Which companies are approaching a cyclical trough, where industry supply is declining, and management points to a demand recovery?
- Create a summary table laying out takeover defenses for all software companies below $1B of market cap.
- Find 100 companies above $2B of market cap that have clearly articulated some benefit from using AI internally as part of their business operations.
Start With Pre-Built Context for Complex Agent Work
To report on “long-term trends driving utility earnings”, agents can begin with CapRelay's prepared industry maps, earnings reviews, and financial history, instead of rebuilding years of context from raw filings and transcripts.
Get better results by spending expensive tokens on analysis, not reconstructing context.
Reduce Agent Errors with Reliable Financial Tables
If you're wondering why your agents are making mistakes when returning financial data, it's because SEC tables are built for human eyes, not agents. CapRelay reconstructs the table and returns structured TSV for AIs.
processor →
```tsv Metric FY26 FY25 Revenue $62,166 $50,000 Operating loss (638) (412) Operating margin (1.0%) (0.8%) ```
Trusted by 60+ institutions managing hundreds of billions
Simple, Transparent Pricing
Priced like a single expert call.
Prices exclude applicable taxes. Sales tax will be added where required.
About CapRelay
CapRelay is NYC-based and was founded by Luyi Zhao and Chi Song.
Luyi and Chi met as analysts at Goldman Sachs, where they sat next to each other for nearly two years working on public and private investments.
Both Luyi and Chi understand the day-to-day work of finance, and how to use the latest technology to help finance professionals find new ideas and save time.
- Tegus, Director of Product Management
- BamSEC, Co-Founder
- Goldman Sachs, Associate, Special Situations Group - Multi-Strategy Investing