Overview
Research Mode executes complex multi-step research tasks using chain-of-thought reasoning and automated tooling. It gathers information from internal data stores and web sources, synthesizes findings, and delivers comprehensive reports, emails, or summaries—all from a single natural language query. Ideal for market research, competitive analysis, trend monitoring, and report generation.
Learning Objectives:
Access and activate Research Mode from chat interface
Formulate effective research queries with clear objectives
Understand chain-of-thought research process and monitoring
Review results with citations from internal and external sources
Repurpose research into actionable formats (reports, emails, summaries)
Prerequisites
You'll Need:
Active TeamAI account
Research question or topic to investigate
Internal data stores uploaded (optional but recommended for proprietary data)
Basic understanding of desired output format
Access Research Mode
Activate Research Environment
Navigate to main chat interface
In left panel, click "Chats"
New chat window opens
Open mode selector
Locate mode options at bottom of chat interface
Select Research Mode
Click "Research" to activate
System configures appropriate model and enables research tools automatically
Result: Chat is now in Research Mode, optimized for multi-step information gathering.
Note: Research Mode may appear as "Research" or with a specific icon in mode selector.
Formulate Research Query
Write Effective Research Request
Be specific about topic
Include keywords and scope
Example: "Tesla stock performance last 12 months" not just "Tesla"
Specify output format
State desired format clearly
Options: "detailed report," "email summary," "executive summary," "bullet points"
Include time frames
When relevant: "past week," "last quarter," "2024 trends"
Send query
Type complete request in chat input
Example: "Research today's AI news and give me a detailed report and email for my boss"
Press Enter to execute
Result: Research Mode begins structured, multi-step research process.
Tip: Research Mode works best for complex queries requiring multiple searches. For simple lookups, use standard chat.
Monitor Research Process
Observe Chain-of-Thought Execution
Watch research plan creation
TeamAI displays plan with numbered steps
Shows what information will be gathered and from where
Monitor step execution
Each step executes sequentially
System shows current action: "Searching internal data," "Searching web," "Analyzing results"
Transparent process allows you to verify approach
Review internal data check
First priority: TeamAI searches your uploaded data stores
Identifies relevant internal documents
Result: Proprietary information is consulted first, saving time and providing relevant context
Observe external search
System transitions to web searches
Performs multiple targeted searches
Gathers information from diverse sources
Result: You see exactly how research is conducted, building confidence in results.
Review Research Results
Evaluate Deliverables
Review comprehensive report
Main content appears in chat
Well-organized sections with clear headings
Logical flow from introduction to conclusions
Verify citations
Source URLs included throughout
Links to internal documents and web pages
Citations allow fact-checking and deeper exploration
Use formatted content
If requested, email/report is generated
Example: "Email to boss" is formatted professionally
Tip: Copy formatted content directly for immediate use
Ask follow-up questions (optional)
Continue conversation to clarify points
Request additional details on specific findings
Refine results without starting new search
Result: You receive actionable, well-sourced research in requested format.
Best Practices
Be specific - Define topic and output format clearly
Include time frames - Add recency: "past week," "12 months"
Use for complex queries - Multi-step research vs. simple lookups
Review research plan - Understand AI's approach to your question
Verify critical data - Click citations for high-stakes decisions
Follow up - Ask clarifying questions in same chat
Leverage internal data - Upload documents for proprietary context
Try different formats - Request reports, emails, summaries based on audience
Troubleshooting
Research Taking Too Long
Cause: Complex query requiring many searches
Solution: Narrow focus. Instead of "AI news," try "OpenAI model releases past week"
Limited Results
Cause: Obscure topic or very recent events
Solution: Broaden search terms or remove time constraints. Check internal data stores are uploaded.
Irrelevant Information
Cause: Query too broad
Solution: Be more specific. Include keywords and context about intended audience.
Common Questions
Q: How does Research Mode differ from standard chat?
A: Uses chain-of-thought reasoning and automated tooling to break complex tasks into steps, search multiple sources, and synthesize information systematically.
Q: Does it access internal documents?
A: Yes. It checks your uploaded data stores first, then supplements with web searches.
Q: How detailed should my query be?
A: State what you need and desired format. Research Mode develops its own detailed plan.
Q: How long does research take?
A: Longer than standard queries due to multi-step process. Time varies by complexity.
Q: Does it provide citations?
A: Yes. Includes source URLs for verification and deeper exploration.
Q: Can it handle specialized topics?
A: Yes, quality depends on available information online and in your internal documents.