Outputs

Learn how Lens delivers synthesized intelligence in formats that fit your workflow.

Structured Research Data

The final stage of the Lens flow is rendering the processed data into a usable format. Every research result can be exported as a raw data dump or a human-readable report.

JSON

Full technical data including source metadata, authority scores, and processed snippets. Ideal for API integration.

Markdown

A beautifully formatted research report with citations, summaries, and links. Ideal for documentation and research notes.

Citations & Provenance

Lens prioritizes transparency. Every synthesized statement in an output is backed by a verifiable source link and a provenance score.

markdown

# Example Markdown Output excerpt
## Summary of fine-tuning techniques
LoRA (Low-Rank Adaptation) significantly reduces the number of 
trainable parameters [1][3]. 

Sources:
[1] arXiv:2106.09685
[3] GitHub: microsoft/LoRA

Exporting Data

Use the CLI or SDK to trigger an export to your desired format.

bash

# CLI Export
lens query "AI Ethics" --format=markdown --output=report.md

# SDK Export
result = lens.query("...")
result.to_json("data.json")

Custom Exporters

Lens is extensible. You can write custom exporter classes to send research findings directly to Notion, Obsidian, or a specialized vector database.

Pipelines