Job Types
This page documents all job types in the system, including both LLM and system jobs.
Analyze Article
analyze_articleLLMGenerates a brief gist (1-2 sentences), creates a detailed summary, and classifies the article into one or more types (earnings, M&A, analyst rating, ...
Analyze Sentiment
analyze_sentimentLLMAnalyzes the sentiment (positive/neutral/negative) toward a specific company mentioned in an article or snippet. Scores on 0-10 scale and generates a ...
Clean Markdown Article
clean_markdown_articleLLMIdentifies promotional paragraphs (See Also links, newsletter CTAs, image attributions) that should be removed from article markdown. Returns paragrap...
Clean Raw Article
clean_raw_articleSystemConverts raw HTML article content to cleaned markdown format. Performs regex-based cleaning and HTML-to-markdown conversion without using LLM.
Extract Entities
extract_entitiesLLMIdentifies companies mentioned in article or snippet content using function calling for database validation.
Generate Story Title
generate_story_titleLLMGenerates a unified professional headline from multiple snippet titles covering the same Google News story.
Generate Talking Points
generate_talking_pointsLLMGenerates AI-powered bull and bear investment arguments for companies using pre-fetched data context (articles, ratings, trades, metrics).
Link Article to Stories
link_article_to_storiesLLMMatches an article to active Google News stories using strict matching criteria.
Link Story to Articles
link_story_to_articlesLLMMatches articles to Google News stories using strict matching criteria.
Score Story Relevance
score_story_relevanceLLMScores a Google News story's business relevance on a 1-10 scale with explanation.
Scrape Article Content
scrape_article_contentSystemFetches full article content from supported sources (Yahoo Finance) using ScrapingBee. Resolves Google News redirect URLs and handles paywalls gracefu...