> ## 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.

# Agent Mode

> Talk to our software to get straight to desired outcomes

Agent Mode offers users an interface to automate major chunks of the administrative work that goes into a round process. In comparison with generic AI models, Agent Mode understands the fundraising context in greater depth and delivers sharper outcomes.

***

## Common Use Cases

The below is a set of common use cases for Agent Mode:

> [card] Discover and Rank Investors
>
> Agent can identify the "most likely" investors with deep reasoning and data points, or qualify an existing list.

> [card] Email Scan
>
> Agent scans your email to pull all investors you have engaged with sharp notes on where each conversation ended.

> [card] Assess Meeting Patterns
>
> Agent scans all prior meeting notes and transcripts to identify patterns into what is working well versus what is not.

> [card] Setup and Manage Investor Pipeline
>
> Agent can create pipelines, add investors, save research, add notes and organise next steps.

With Metal, founders are able to consolidate the full company and round context, and equip Agent Mode to adapt to the evolving situation within their raise.

> [!TIP]
>
> **Designed to Deliver on Outcomes**
>
> By design, Agent Mode is equipped to deliver on outcomes that historically required founders to spend months of research consolidation in spreadsheets.

***

## User Context for Agent Mode

Agent Mode is equipped to understand your raise process and adapt to rapid evolutions in your round context. The below set of data sources automatically feed the platform with fresh context on your round and company:

> [card] Calendar Data
>
> Provides input on who you're meeting with.

> [card] Meeting Intelligence
>
> Data on feedback from investor conversations.

> [card] Email Data
>
> Insight into how investor engagements progress.

> [card] Relationship Context
>
> How your network connects into VCs.

> [card] Pipeline Data
>
> Data on pipeline velocity and investor targeting.

> [card] Round Dynamics
>
> Context on lead versus party rounds.

> [card] Existing Investors
>
> List of investors that are on your cap table.

> [card] Round Stage
>
> Stage at which you're raising.

> [card] Pipeline Benchmarks
>
> 1st to 2nd call conversion relative to benchmarks.

With a data dense environment, the platform is able to keep up with key nuances of how your round progresses. In addition to the above ongoing data sources, Agent Mode also has context on your pitch deck, existing investors,

***

## Intelligence Accessible to Agent Mode

Equipped with off-market intelligence, Agent Mode has access to a broad set of data points that are used selectively to deliver sharp outcomes.

> [card] Investor Memory Layer
>
> Access to 1m+ data points on venture investors through industry APIs, public content and off-market platform data.

> [card] Deal Memory Layer
>
> In-depth deal memory layer that brings intelligence on deal processes for specific VC firms.

> [card] Scenario Based Intelligence
>
> Insights from thousands of raise processes on scenario-based actions and approaches that worked well.

> [card] Relationship Context
>
> Industry intelligence on key relationships across founders and VC firms, including portfolio and co-investing activities.

With access to rich sources of intelligence and off-market data, Agent Mode is capable of delivering much sharper responses and outcomes than that of generic models.

***

## Multi Step Prompts to Try

For founders looking to test out the effectiveness of Agent Mode, we recommend the following set of common use case prompts;

Build a targeted investor pipeline

Find 20 seed investors that fit our round, prioritise those where we have strong introduction paths, and add the best 10 to a new pipeline.

Research and qualify investors

Find active Series A investors in the US and Europe that invest in B2B AI companies. Qualify them based on fit, recent activity, similar investments and ability to lead.

Prepare for upcoming meetings

Review my upcoming investor meetings, prepare a brief for each one, and tell me what I should accomplish in every conversation.

Prepare investor follow-ups

Find investor conversations where I have not received a reply in the last seven days and draft a contextual follow-up for each one.

> [!TIP]
>
> Agent Mode works best when you describe the desired outcome and explicitly mention important constraints.

***


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