{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/c26fff8ab35f409997863479564c958e\" frameborder=\"0\" width=\"1920\" height=\"1440\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1440,"width":1920,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1440,"thumbnail_width":1920,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/c26fff8ab35f409997863479564c958e-3f3f8a51cbe1133e.gif","duration":307.079,"title":"Samra Multi Agent Stock Research Workflow","description":"This Loom introduces how Stock Market Research Assistant uses routing and guardrails to answer questions safely. It shows a workflow where Apple’s closing price for last week is retrieved via a safe select-only SQL query from Postgres, including currency to avoid silent USD-INR mix-ups. It then covers filing retrieval using Pinecone with hybrid page-level citations from actual 10-K PDFs and a groundedness check to prevent hallucinated numbers. For live news, it switches to a web agent using timely sources, and a router runs SQL and filing data together to synthesize one coherent answer while blocking prompt injection and SQL abuse."}