> ## Documentation Index
> Fetch the complete documentation index at: https://docs.metal.so/llms.txt
> Use this file to discover all available pages before exploring further.

# Investor Memory Layer

***

With access to millions of public and off-market data points, Metal has built an investor memory layer that deeply understands the investing patterns, inclinations and behaviours of venture investors.

***

## Underlying Data Sources

Investor memory draws on data from a broad set of sources explained below. For data ingested via each source, the platform has strong QA processes built in to ensure a continuous self-improvement mechanism.

**Industry APIs**

By partnering with leading industry APIs, Metal has access to structured data on investment rounds from the past 15 years. With each API partner, the platform pulls data that is verifiable and reliable, and avoids using data points that are not particular strengths for that partner.

**Public Content**

By gathering data from a broad set of sources across the web, Metal has amassed large data sets that capture investor sentiment and inclinations. Further, by sourcing industry records from public sources, the platform relies on a rigorous information verification system built in-house.

**Tribal Knowledge**

On a quarterly basis, Metal uses tribal knowledge into structured data through first party interviews and collection methodologies. Customers using the platform frequently contribute their views and facts about the investors they speak with.

**Proprietary Data**

Using data generated on the platform, Metal has access to off-market data that drives a layer of proprietary intelligence on the venture industry. Proprietary data is stored in a manner that is secure and anonymised.

By accessing data from a broad mix of sources, the platform is able to gather accurate data and use that to build a nuanced understanding of venture investors. With each passing quarter, as the proprietary intelligence grows, the platform experience improves automatically for all users.

***

## Structured Insight via the Memory Layer

Equipped with proprietary intelligence on venture investors, the platform draws on large data sets to build structured insight into the investing patterns, inclinations and behaviours of venture investors. For institutional investors, this is managed separately for investments firms and for partners.

For most VC firms, the platform holds structured insight into what they seek:

> **Business Models**
>
> Business models or market theses is the firm interested in or excited about.

> **Founder Backgrounds**
>
> An understanding of the types of founder backgrounds that the firm backs.

> **Process Insights**
>
> Insights on the investment process for a given firm (time to decision).

> **Key Patterns**
>
> Specific patterns that emerge in their recent investments.

The above are a few examples of the the of structured insight that comprises the investor memory layer.

***

## Application of the Investor Memory

Across the platform, the investor memory layer brings workflow intelligence to the following capabilities:

> **Pre Meeting Briefs**
>
> Intelligence on likely questions and focus areas for an upcoming investor call.

> **Post Meeting Intelligence**
>
> Intelligence on what went well versus can be improved after an actual investor call.

> **AI Mock Calls**
>
> AI-led mock calls that mimic the investing inclinations of the selected investor.

> **Data Rooms**
>
> Insight into what needs to improve for pitch decks or round collateral.

> **Relevance Ranks**
>
> Relevance ranking of investors based on investor and user memory layers.

> **Relationship Intelligence**
>
> Insight into connections that are connected with a given investor.

> **Investor Updates**
>
> Usage of investor memory to recommend improvements to investor updates.

> **Investor Pipeline**
>
> Proactive recommendations for new tasks or investor additions to the pipeline.

With proprietary data driving workflow intelligence, the platform is able to enable founders in novel ways with insights on how the round is progressing, or with suggestions on areas that need to be re-evaluated.

***


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.