Objective

The purpose of this article is to provide direction to Outreach Admins in setting up and managing Research Agent. The Research Agent feature provides AI-powered research automation to Outreach Users by extracting insights from internal or external sources and saving them into account fields for use in personalization, planning, segmentation, or automation.

Applies To

• Outreach Amplify Package
• AI Content > Research Agents
• Account Object
• Personas: RevOps (setup and execution), AE/SDR (consumption only)

Before You Begin

• Research Agent must be configured by RevOps users from the Research Agents tab in the Outreach platform
• At launch, only the Account object is supported
• Each Research Agent must use either internal or external sources (not both)
• Running agents is permission-controlled
• Results are stored in custom fields on accounts
• All runs are logged and tracked

Overview

What is Research Agent?

Research Agent is a tool that automates account-level research using AI. It extracts structured insights from either internal (Kaia call transcripts, Kaia meeting summaries, emails) or external (Gemini-powered search and Google-grounded web content) sources. The research results are saved into account fields and used across Outreach — in smart account plans, reporting, filters, and triggers.

Configuration

RevOps: Set Up and Manage Research Agents

Go to AI Content → Research Agents. From this list view, RevOps can:

• View existing agents
• Create new ones
• Test and refine agent configuration in Playground Mode

To create a new Research Agent:

1. Click New Research Agent
2. Choose a blueprint or start from scratch
3. Name the agent and add a description
4. Confirm object: Account (pre-set)
5. Select the account field where research results will be stored
6. Choose one source type:
   1. Internal: Kaia calls, Kaia meeting summaries, and emails
   2. External: Gemini + optional Google Search
7. Use Playground Mode to preview and iterate on outputs

Once saved, the agent is available for use across the platform.

Running Research Agents (RevOps, AE, SDR based on access)

Option 1: Account List View (Bulk Run)

1. Go to the Account List View
2. Use Manage Columns to add RA fields
3. Select accounts
4. Click Research AI Fields
5. Choose fields to run from the pop-up
6. Execute the run. Populated fields will be updated unless override is enabled

Option 2: Account Plan View (Manual Run)

Step1: Setup Research agent tiles in Account Plan

1. Add the Research Agent tile to the Prospect or Account Overview layout
2. Click Configure to select which Research Agent fields to surface in the tile

Step2: Run Research Agents

1. Click “Research AI Fields” to launch the same field selection pop-up
2. Once executed, selected fields auto-populate in the plan tiles

Step3: Utilize Research Agent Output in Research agent tiles

1. Use the “copy” icon for quick sharing of insights and pasting content to the documents of your choice
2. Use the “info” icon to view detailed citations for each field

Enhancements for Context & Coverage of Internal Research

1. Buyer–Seller Tagging

Improves contextual accuracy by enabling the LLM to distinguish between buyer and seller voices during Kaia calls and meetings — resulting in sharper, buyer-focused summaries.

2. Expanded Coverage

Research Agent now processes a larger data window:

• Emails: Increased from 80 → 500
• Meetings/Calls: Increased from 10 → 80 

This gives the agent significantly more context for analysis and insight generation.

3. Time-Frame Filters

Users can now define the time period for their research, choosing from preset options or custom ranges:

• All time
• last 7/30/90/365 days
• 7/30/90/365 days prior to last conversation
• Custom: Choose any custom date range

In-Product Guidance

Collapsible “How Research Agent Works” Banner

To help users better understand what Research Agent can and cannot do, a new collapsible banner has been added to the configuration screen. This banner:

• Explains the two modes of operation:
  • Web: Performs internet searches to fetch the most up-to-date online information.
  • Outreach (Emails, Kaia Calls/Meetings): Searches internal data like meeting transcripts and emails.
• Clarifies the expected scope and limitations of Research Agent.
• Can be expanded or collapsed as needed, keeping the UI clean and informative.

Research Agent Run Notifications

• Users receive in-app notifications when Research Agents are executed, showing the status of the run and field updates.
• Notification settings can be configured under Personal Settings → Notifications → Research Agent Run Status to enable or disable visibility across in-app, email, or Slack channels.
• These alerts provide transparency and feedback, especially during bulk or plan-based research actions.

.         

Governance

Global Toggle in Org Settings 

A new Global Research Agent toggle is available in the Gen AI Settings page under Org Settings.

• This feature is enabled by default, allowing admins to turn off Research Agent if not required for their organization.
• Provides centralized control for system administrators to manage Research Agent availability across all users.

Profile Access Governance

Admins can control access to Research Agents through granular governance settings that determine what actions each profile can take and on which agents. These settings ensure data security, accountability, and operational control.

Access Levels: 
Admins can configure Read, Update, and Delete permissions across different scopes:

• None
• Owned by them
• Owned by them and their reports
• Owned by them, their peers, and their reports
• Owned by anyone

Each permission (Read, Update, Delete) can be independently scoped based on organizational needs.

Additional Permissions:

• Can create: Allows users in this profile to create new Research Agents.
• Can run research agents: Allows users to execute or trigger research agents.

These governance controls allow organizations to tailor access based on user roles, ensuring that only the right individuals can view, modify, or remove Research Agents.

Key Use Cases

Internal Research (via Kaia calls, meetings, and emails)

Handoff & Onboarding Support 
Surface success criteria, implementation details, and use cases from past conversations to ensure smooth transitions from Sales to Success.

Churn Risk Identification 
Analyze drop-offs in engagement and stakeholder activity to detect early churn signals and proactively retain accounts.

Competitive Intelligence 
Identify competitor mentions from prior interactions to prepare targeted messaging and objection handling.

External Research (via Gemini + Search)

Company Summary & Contextual Insight 
Aggregate public information about business goals, challenges, and leadership priorities to power relevant and contextual outreach.

ICP Fit & Qualification 
Evaluate firmographic and technographic fit to prioritize accounts effectively during prospecting or segmentation.

Frequently Asked Questions

What is Research Agent? 
An AI-powered tool that extracts account-level insights from either internal or external data and saves them into Outreach fields for scalable use.

What sources can be used? 
Each agent supports one of the following:

• Internal: Kaia calls, Kaia meeting summaries, and account emails
• External: Gemini-based web search with optional Google grounding 
  Only one source type can be used per agent.

What Outreach plan includes this feature? 
Outreach Amplify SKU. It will be auto enabled for customers on this plan.

Where do I set it up? 
Go to AI Content → Research Agents to manage agents. Only RevOps or Admin-level users can configure and deploy agents.

What objects are supported? 
Currently, only Account is supported.

Who can run Research Agents? 
Any user with access at Admin discretion. Runs can be initiated from Account List View or Account Plan View.

Can I test an agent before rollout? 
Yes. Playground Mode allows RevOps to preview results on specific accounts and refine configuration.

Where is output stored? 
In a designated custom field on the account, chosen during agent setup.

Can I run in bulk? 
Yes. Select multiple accounts in list view, launch the Research AI Fields tool, and run the agent on selected fields.

Can I rerun an agent? 
Yes. Admins can configure whether the agent should skip already-filled fields or override them with fresh output.

How can I view citations for Research Agent output?  
Each Research Agent tile includes an info icon next to the copy icon. Click this icon to view citations for each field displayed in the tile. This provides visibility into data sources and ensures output transparency. 

What happens if a run fails? 
Failures are logged and tagged with an error type (e.g., missing input, failed search). RevOps can access logs for debugging. User can view RA run status with below ways -

• System Activity -> Bulk Actions
• In-App notification

What types of fields are supported?

• Plain text (for filtering, sorting)
• Formatted text with citations and hyperlinks

Can I see the source of insights? 
Yes, depending on the type of research -

• Internal Search -> Yes
  • Internal sources include quoted snippets from Kaia transcripts, emails, or meetings
• External -> No
• External + Websearch -> Yes
  • Web sources include URLs (e.g., company websites or Crunchbase)

Note: Any Posts/Questions/Comments that contain PII will likely be removed to ensure user safety, security, and privacy in the forum.

Is usage tracked? 
Yes. Research runs are logged and tracked, including whether external search was used. Usage is visible in dashboards and admin reports.