Monitor daily stock sentiment, analyze news with Gemini, and log results to Sheets to inform trades at market open.
The AI agent reads a Google Sheet with tickers, fetches the latest headlines from EODHD, and analyzes them using Google Gemini to produce a sentiment score and rationale. It runs automatically each morning, delivering fresh insights before the market open. It logs results to Google Sheets for historical tracking and easy comparison across stocks.
Concrete actions the AI agent performs to generate sentiment insights.
Read tickers from the Google Sheet.
Fetch the latest news for each ticker from EODHD.
Analyze the news with Gemini to generate a sentiment score and rationale.
Validate the AI output JSON for consistency.
Log the score, rationale, ticker, and date to Google Sheets.
Flag and correct invalid tickers or formatting issues.
Before → manual data collection caused delays and inconsistent inputs. After → the AI agent automates tickers, news aggregation, analysis, and logging with auditable results.
A simple 3-step flow for non-technical users.
The AI agent triggers daily at a configured time and reads ticker symbols from the Google Sheet.
For each ticker, it calls the EODHD API to fetch latest articles and uses Gemini via LangChain to generate a sentiment score and rationale.
The AI agent validates the JSON output and appends the score, rationale, ticker, and date to Google Sheets.
One realistic scenario demonstrating task, time, and outcome.
Scenario: At 8:00 AM, the AI agent reads a 10-stock watchlist from Sheets, fetches the latest headlines via EODHD for each ticker, and uses Gemini to generate sentiment scores (-1 to 1) with rationales. The results are validated, formatted as JSON, and appended to the same Sheets with the current date. By 8:05 AM, the sentiment snapshot is ready for review and trading decisions.
One supporting sentence.
Need quick, structured sentiment signals to guide small portfolios.
Want pre-market sentiment to decide intraday trades.
Seek consistent signals across many holdings with auditable data.
Require reproducible inputs for research notes and reports.
Need scalable sentiment extraction for multiple sectors.
Desire data-driven signals that feed into models and dashboards.
One supporting sentence with short explanation.
Reads tickers from sheets and writes results back, enabling easy monitoring and historical tracking.
Fetches the latest stock news for each ticker to feed sentiment analysis.
Coordinates prompts, parsing, and data flow between data sources and Gemini.
Performs sentiment analysis and generates rationale for each ticker.
Parses and validates the AI output JSON to ensure consistent logging.
One supporting sentence with short explanation.
One supporting sentence with short explanation.
The AI agent pulls headlines from EODHD for each tracked ticker and analyzes them with Google Gemini. Inputs are read from and logged back to Google Sheets to provide a complete audit trail. The sentiment scores range from -1 (negative) to 1 (positive), with a detailed rationale included to explain the scoring. Outputs are validated as JSON to prevent parsing errors downstream.
Yes. You maintain a Google Sheet with a list of tickers. The AI agent reads this list at the scheduled time and processes each ticker in sequence, allowing you to add or remove symbols as your strategy evolves. Validation checks prevent invalid tickers from skewing results, and errors are surfaced for correction.
If the EODHD API or Gemini access encounters an issue, the AI agent records the error alongside the attempted ticker, and the schedule continues. You receive a clear log entry and a notification path can be configured. This ensures a partial result with explicit failure notes rather than silent data gaps.
The agent is designed for a daily cycle, but you can modify the trigger to run additional times as needed. Each run maintains separate date-stamped results in Sheets for accurate historical tracking. However, overlapping runs should be managed to prevent duplicate entries.
Yes. The sentiment score is standardized on a -1 to 1 scale, with -1 meaning strongly negative and 1 meaning strongly positive. The rationale aligns with the underlying articles and a transparent scoring rubric. The output is consistently structured to facilitate automated logging and analysis.
Scores, rationales, tickers, and dates are appended to Google Sheets in a structured format. The JSON output is validated before logging to prevent malformed data. Historical entries enable trend analysis and reporting across time.
The current design targets US equities with English-language news from EODHD. Extending to additional markets or languages would require source configuration and language model prompts adjusted for multilingual sentiment analysis. If you plan expansion, the AI agent can be updated to handle multiple data sources and localized prompts.
Monitor daily stock sentiment, analyze news with Gemini, and log results to Sheets to inform trades at market open.
