Documents and PDFs
Research, interviews, contracts, specifications, and scanned files.
Multi-file analysis report
Cross-analyze multiple files and produce an evidence-backed report
Put PDFs, documents, spreadsheets, images, and web sources in one workspace. Flowmint extracts content, aligns definitions, compares evidence, and produces reviewable reports, data tables, or presentations.
ANALYSIS REPORT
Where first-use activation breaks down
Evidence combined from interviews, feedback, and behavioral data.
01
Setup load
02
Value delay
03
Choice overload
Evidence strength
Setup load
Value delay
Choice overload
Bring different source formats together
Add only what helps the agent understand the target and verify the result.
Spreadsheets and data
Metrics, records, categories, and time series in XLSX or CSV.
Images and web sources
Screenshots, charts, public pages, and links that must remain traceable.
Deliverables
Deliver conclusions together with their evidence
Cross-file findings
Merge repeated themes and surface differences, anomalies, and conflicts.
Evidence index
Keep references to source files, tables, and web pages for verification.
Structured results
Create classifications, calculations, priorities, and reusable data files.
Shareable deliverables
Produce Markdown, DOCX, PPTX, XLSX, or a previewable report page.
Process
Analyze around the question instead of summarizing files one by one
01
Inventory the sources
Identify formats, time ranges, definitions, and missing information.
02
Extract and organize
Read text, tables, and images into a shared analytical framework.
03
Analyze and verify
Calculate metrics, compare sources, and separate facts from inference.
04
Create deliverables
Package findings, evidence, and next actions into editable files.
Example workflow
Combine customer feedback, interviews, and usage data into priorities
See how the agent handles mixed sources, preserves evidence, and creates a report and priority table.
View the full example“Analyze these interview notes, customer feedback files, and usage data. Find repeated issues and conflicting evidence, prioritize improvements, and create a report I can share.”
Multi-file analysis questions
How many files can it analyze at once?
It depends on file size, model context, and task complexity. The workspace lets the agent extract in batches and save intermediate results instead of forcing every source into one prompt.
Can it read scans and text inside images?
Yes. Vision-capable models can inspect images directly, while OCR is available for text extraction. Important tables and complex layouts should still be verified.
How do I reduce unsupported conclusions?
Ask the report to cite a source file and evidence for each important finding and to separate facts, inferences, and unresolved questions.
Can I change the analysis framework later?
Yes. Sources, intermediate files, and outputs stay in the workspace, so you can revise categories, metrics, or audience without starting over.
Bring the source material into one workspace
Describe the question first, then upload the files and sources.