Starting a conversation with Ivy
To open the chat panel, click the Ivy icon in the upper right hand corner of the dashboard. You can also type directly in the chat panel, and press send to enter the primary Ivy chat interface. Ivy is designed to handle a broad range of trading-related questions across multiple dimensions of the market:Earnings & Fundamentals
Ask about upcoming earnings dates, analyst estimates, historical beat/miss rates, and key risks heading into a report.
Sector & Macro Outlook
Get a read on sector rotation, macro headwinds, or how a Fed decision might ripple through specific industries.
What's Moving Today
Ask Ivy to surface the top movers, unusual volume, or breaking catalysts across the market right now.
Portfolio Questions
Explore concentration risk, correlated positions, or how a new trade might fit within your existing holdings.
Example prompts to get started
Try asking Ivy questions like:- “What are the key risks for TSLA going into earnings?”
- “Which sectors are showing relative strength this week?”
- “What’s moving in biotech today and why?”
- “How correlated are my current positions?”
- “Give me a bull and bear case for NVDA at current levels.”
Research agents
Beyond single-turn Q&A, Ivy can launch research agents — autonomous multi-step workflows that gather, synthesize, and summarize information across several sources before returning a consolidated answer. Where a standard chat response might draw on a single data source, a research agent might simultaneously pull SEC filings, recent news headlines, earnings call transcripts, and analyst commentary, then stitch them into a coherent narrative.When to use a research agent
Research agents are most useful when your question requires breadth — for example:- “Give me a full pre-earnings briefing on AMZN: recent guidance, analyst sentiment, and top risks.”
- “Summarize the last three 10-Q filings for META and flag any changes in forward-looking language.”
- “Scan the news for anything that could move energy stocks this week.”
How to trigger a research agent
When Ivy detects that your question warrants a multi-source deep dive, it will automatically launch an agent and show a progress indicator while it works. You can also explicitly invoke an agent by opening the agents module, activating an agent, and prefixing your prompt:- “Run a research deep-dive on MSFT ahead of next week’s earnings.”
- “Agent: summarize recent FDA decisions that could impact MRNA.”
Understanding Ivy’s responses
Ivy structures its responses to be scannable and actionable. Long-form analysis is broken into labeled sections; lists of catalysts or risks are bulleted for quick review; data points (price targets, earnings estimates, IV percentiles) are called out inline. A few important things to keep in mind as you work with Ivy:- Ivy is AI-generated. Responses synthesize available data, but they can contain errors, omissions, or outdated information. Always cross-reference critical data points with primary sources.
- Context matters. The more specific your question, the more precise Ivy’s response. Vague prompts produce general answers; specific prompts with a ticker, timeframe, and angle produce targeted analysis.
- Session memory is scoped. Ivy remembers the thread of your current session. Starting a new chat clears the prior context.