Get answers from your files
Analyze one or more selected workspace files, extract verified findings per file, and synthesize a cited answer with clickable references.
Created by Chris Moen • Version 65 • 13 steps
What you get
- Analyze one or more selected workspace files, extract verified findings per file, and synthesize a cited answer with clickable references.
Integrations
- openai
- mistral
- ai
- ocr-api
- Ocr Api
- Document Ocr Provider
How it works
- 13 workflow steps run in Breyta to produce the app result.
Analyze selected files with grounded citations
Ask a question about one or more workspace files or saved results and get a synthesized answer backed by source-level evidence.
This flow analyzes the materials you choose for each run, extracts verified findings per file, and returns a final answer with clickable references.
What you can use it for
- Compare themes, requests, risks, and decisions across multiple sources
- Review interview transcripts, meeting notes, research docs, and saved results
- Turn a set of files into one grounded answer instead of reading them one by one
What setup looks like
- Setup is minimal
- No files are locked in at install time
What each run looks like
- Enter the question you want answered
- Select the workspace files or saved results you want analyzed
Tips for better analysis
- Choose only the most relevant files for the question
- Ask focused questions when you want sharper conclusions
- Use saved results when you want to analyze output from another workflow
Good to know
This flow works best with text-based materials such as transcripts, notes, documents, CSV files, JSON, and other structured workspace resources.
FAQ
What does the Get answers from your files app do?
This app analyses multiple workspace files to find specific information, extract verified facts, and create a single answer based on those findings.
How does the app process my documents?
The app examines each file individually to pull out relevant data points. It then combines these insights into a final response that includes clickable references and citations for every claim.
Which AI services and integrations does the app use?
The workflow uses Optical Character Recognition API for reading images and documents, alongside models from OpenAI and Mistral to perform the deep analysis.
How do I set up and customise this workflow?
You can select any files within your workspace to start the analysis. The app handles research tasks automatically, so you don't need to configure complex steps to get your cited answers.