​

How the AI Research Assistant Works

The AI Research Assistant is your personal investment research analyst. It answers questions about stocks, markets, and investing by combining advanced language AI with real-time financial data. Here’s how it works under the hood.

​

The Basic Process

When you ask a question, here’s what happens:

1. You ask a question in plain English (e.g., “What are Tesla’s profit margins?”)
2. The AI understands your intent - It figures out what you’re really asking, which data you need, and how to structure a helpful response.
3. It pulls relevant data - The AI queries our financial databases for current and historical data on the stocks or topics you asked about.
4. It synthesizes an answer - Combining the data with its knowledge of finance and investing, the AI writes a clear, useful response.
5. You get a research-backed answer - Not just an opinion, but an analysis grounded in real numbers.

​

What Makes This Different from ChatGPT

You might wonder: why not just ask ChatGPT about stocks? In short: Rallies AI knows what’s happening in markets right now, not just what it learned during training.

​

Data Sources

The AI pulls from multiple professional financial data sources to answer your questions: Market Data

• Real-time and delayed stock prices
• Historical price data
• Trading volume
• Market cap and shares outstanding

Fundamental Data

• Income statements
• Balance sheets
• Cash flow statements
• Key financial ratios (P/E, margins, ROE, etc.)

Analyst Information

• Analyst ratings and price targets
• Earnings estimates
• Recommendation changes

Corporate Events

• Earnings reports and transcripts
• SEC filings (10-K, 10-Q, 8-K, etc.)
• Dividend history
• Stock splits

Ownership Data

• Institutional holdings (13F filings)
• Insider transactions
• Politician trades (from public disclosures)

News & Sentiment

• Recent news headlines
• Market-moving events

Learn more about our data sources →

​

Conversation Memory

The AI remembers what you’ve discussed within a conversation. This means you can:

• Ask follow-up questions without repeating context
• Drill deeper into topics (“Tell me more about that”)
• Make comparisons to previous points (“How does that compare?”)

Example conversation:

You: What are Apple’s profit margins? AI: [Provides Apple’s gross and net margins with historical context] You: How does that compare to Microsoft? AI: [Compares both companies without you needing to re-explain] You: Which has grown margins faster? AI: [Analyzes margin growth trends for both]

Starting a new conversation clears this context. If you want to reference something from a previous chat, you’ll need to provide that context again.

​

Advisor Lenses

The AI can analyze stocks from different investment perspectives using “Advisor Lenses”: You can ask the AI to analyze a stock through a specific lens:

• “Analyze Apple from a value perspective”
• “What would a growth investor think about this stock?”
• “Give me the quality case for Microsoft”

​

Deep Research Mode (Pro)

For comprehensive analysis, Pro users can activate Deep Research Mode. This runs 100+ analysis steps in about 60 seconds, covering:

• Financial statement analysis
• Valuation metrics and comparisons
• Technical analysis
• Analyst sentiment
• Insider and institutional activity
• News and recent developments
• Bull and bear cases

To use it, just ask: “Deep research on [ticker]” or “Give me a comprehensive analysis of [company]” Learn more about Deep Research →

​

What the AI Can and Can’t Do

The AI can:

• Analyze financial data and metrics
• Explain investing concepts clearly
• Compare companies on various factors
• Summarize earnings reports and SEC filings
• Show you what analysts think
• Identify trends in data
• Help you understand your portfolio
• Answer educational questions about investing

The AI cannot:

• Tell you to buy or sell specific stocks
• Provide personalized investment advice
• Predict future stock prices
• Guarantee investment outcomes
• Know your personal financial situation
• Replace a financial advisor

Learn more about AI limitations →

​

A Note on AI-Generated Content

The AI generates responses based on data and its training. While we work hard to ensure accuracy:

• Always verify important information, especially before making investment decisions
• The AI may occasionally make errors or misinterpret data
• Financial data has inherent delays and potential inaccuracies
• The AI’s analysis reflects patterns in data, not certainty about the future

You can report inaccurate responses using the feedback buttons below any AI response. This helps us improve. How to report issues →

---

Important: Rallies.ai is an investment research and educational platform. We are not a registered investment advisor. The AI provides information and analysis, not personalized advice. Read our full disclaimer →

---

​

Frequently Asked Questions

​

How current is the AI’s information?

Market data is real-time for Pro users and 15-minute delayed for free users. Financial statements update as companies report (typically quarterly). News and filings are incorporated as they’re published. Learn more about data freshness →

​

Does the AI learn from my questions?

Your conversations help improve Rallies over time, but the AI doesn’t build a persistent profile of you across sessions. Each new conversation starts fresh. We use anonymized, aggregated patterns to improve the service.

​

Can I trust the AI’s analysis?

The AI provides data-driven analysis based on public financial information. It’s a research tool to inform your thinking, not a replacement for your own judgment or professional advice. Always verify important information and consider consulting a financial advisor for significant decisions.

​

Why does the AI sometimes refuse to answer?

The AI won’t provide personalized buy/sell recommendations because that would constitute investment advice, which requires regulatory registration. It also avoids speculative predictions about future prices. These limitations exist to keep you safe and keep Rallies compliant with financial regulations.

---

​

Related Articles

• What Can I Ask the AI?
• Getting Better Answers: Prompt Tips
• Understanding AI Limitations
• Where Does the AI Get Its Data?