The research workspace

Search and screen literature, collect interviews and surveys, analyse data, build figures and write the manuscript without separating the evidence from the decisions made around it.

Check availability Explore the workspace

SixSentences_· Research, on the recordWorkflowsp. 02

01 / 07Evidence synthesis
SixSentences_ turns a research question into a review workspace where retrieval, screening decisions, full texts and the final evidence base remain connected.
Explore the workflow
02 / 07Literature discovery
SixSentences_ combines scholarly metadata, the live web, known papers and uploaded sources in one literature search where every result keeps its origin and status.
Explore the workflow
03 / 07Evidence-linked writing
The Manuscript Studio combines a structured writing project, multi-file LaTeX source, rendered PDF, bibliography and an agent that can work against the study record.
Explore the workflow
04 / 07Qualitative research
SixSentences_ supports uploaded interview recordings and AI-led sessions, then keeps transcripts, codes, themes, memos and verified quotes in one study workspace.
Explore the workflow
05 / 07Data Lab
The Data Lab combines versioned tabular data, reproducible analysis steps, charts and a conversational agent that can explain results without inventing values.
Explore the workflow
06 / 07Visual Lab
The Visual Lab plans and renders scientific figures from a brief, then applies a review pass for hierarchy, labels and connections before the researcher approves the output.
Explore the workflow
07 / 07Survey research
SixSentences_ combines conversational survey design with a structured questionnaire builder, participant access controls, response monitoring and research-data analysis.
Explore the workflow

Demo review · run fmvx3f8j2k

Question

Search string · ready for your methods appendix

"large language model"LLM"GPT""generative pre-trained transformer""Claude""Llama"

AND"systematic review""literature screening""title and abstract screening""citation screening""evidence synthesis"

ANDreliabilityagreementsensitivity"recall""work saved over sampling"accuracy

0

Identified

0

Duplicates removed

0

Screened

0

Excluded

0

Unsure

0

Included

Fig. 1The demo review “How reliable are large language models for systematic literature screening?”, run fmvx3f8j2k. Every tile is a real record; which works the 37 duplicates copy is illustrative, their count is real. Hover a paper to read its decision.

Fig. 1SixSentences_ AI literature search showing ranked research papers, source metadata and review controls.

Fig. 1A thesis draft in the Writer's template, grounded on the demo review (run fmvx3f8j2k): main.tex, lines 70 to 76. The log is tectonic's console for this build; the pages are pages 6 and 12 of its PDF.

Fig. 1The study builder of the demo study. The designer chat drafts the guide, its probes and the consent text, and every field stays editable.

Fig. 1The evidence table of the demo review, run xpb4qhfv3z: 30 included studies, 5 fields. Values are verbatim, shortened to fit a cell; a sage dot marks a quote located in the source, an amber ring one that needs a source check. Hover a cell to read it with its study.

Fig. 1The Visual Lab of the demo workspace: figure types, output size, canvas and review passes above, and four finished visuals below.

Fig. 1Question 1 of the survey “Study habits and AI tools”. Every dot is one of its 24 responses, in the order they arrived: 12 daily, 8 weekly, 4 rarely, 0 never. Demo data created by our team, with no real respondents.

SixSentences_· Research, on the recordReferences

Back matterSources

Sources behind the feature previews: team-created demo data, cited papers and the open-source repository. Numbered markers across the site open their entries here.

[1]
Demo review, run fmvx3f8j2k
How reliable are large language models for systematic literature screening? 457 records identified, 37 duplicates removed, 420 screened, 384 excluded, 6 unsure, 30 included. SixSentences demo workspace.
Systematic reviews
[2]
Oami et al. 2025, Research Synthesis Methods
Takehiko Oami, Yohei Okada and Taka-aki Nakada: Optimal large language models to screen citations for systematic reviews. Research Synthesis Methods 16, 2025. © The Author(s), 2025, open access under CC BY 4.0. The quick answer cites its passage on page 7, shown with the app's highlight.
doi:10.1017/rsm.2025.10014 CC BY 4.0 Literature search
[3]
Demo study: How students use AI in everyday studying
Simulated AI-led interview with timestamps and a mock consent step. Two analysis quotes match the team-created demo transcript. No real participants were involved.
Qualitative interviews
[4]
Survey vzvy62phzj: Study habits and AI tools
24 responses to four questions, published into the Data Hub as a dataset. Demo data created by our team, with no real respondents.
Research surveys
[5]
Evidence table, run xpb4qhfv3z
30 included studies x 5 fields, 150 cells: 133 values with a supporting quote, 119 of those located in the source, 17 cells not reported.
Research data analysis
[6]
Sanghera et al. 2025, Journal of the American Medical Informatics Association
High-performance automated abstract screening with large language model ensembles.
doi:10.1093/jamia/ocaf050 Manuscript Studio
[7]
Demo manuscript, AI contribution log
6 requests, 5 proposed edits, 4 applied: 1,901 characters added and 1,745 removed, written into the disclosure paragraph.
Manuscript Studio
[8]
SixSentences community edition
The complete community application stack under Apache 2.0.
github.com/SixSentences/sixsentences Code availability

SixSentences_· Research, on the recordContinue through the workspacep. 03

The research workspace

01Evidence synthesis 02Literature discovery 03Evidence-linked writing 04Qualitative research 05Data Lab 06Visual Lab 07Survey research

Fig. 2The seven workflows. Each line joins a guide to one it continues to; hover a workflow to follow its lines.

Check availability

hello@sixsentences.com Open source on GitHub

One study. One connected record.

The research workspace

Systematic reviews with an audit trail, not a black box.

Search research papers with a trail you can inspect.

Write the manuscript where the evidence already lives.

Interview data that stays connected to the conversation.

Research data with exact analyses and visible provenance.

Scientific figures that are grounded, reviewed and editable.

Build, host and analyse a survey in the same study.

References

Continue through the workspace