Skip to content

01 · PLATFORM — THE CAPABILITY SHEET

Everything Littview does.

The complete reference, organized the way a systematic review actually runs — import to report. Every line on this page ships in the product today; nothing here is a roadmap item.

PREFER THE NARRATIVE? SEE HOW IT WORKS →
39
CAPABILITIES
7
STAGES
0
ROADMAP ITEMS

F01 / IMPORT & ORGANIZE

Bring in the whole search.

6 CAPABILITIES

  • .BIBBIBTEX
  • .RISRIS
  • .TXTPUBMED MEDLINE
  • .XMLENDNOTE
  • ZOTEROPDF LIBRARY
F01.1

Four import formats

BibTeX, RIS, PubMed MEDLINE, and EndNote XML — thousands of references in a single batch.

F01.2

Zotero PDF import

Pull full-text PDFs straight from your Zotero library; references come in via BibTeX export.

F01.3

Batch validation

Malformed entries are flagged with a clear error report — never silently dropped.

F01.4

Duplicate detection

Automatic detection across all your libraries, with an adjustable similarity threshold.

F01.5

Auto-resolve or side-by-side

Bulk-resolve the obvious duplicates; review borderline pairs manually, field by field.

F01.6

Tags & organization

Tag references and keep every source library visible and editable in one place.

F02 / SCREEN

Decide fast without losing rigor.

F02.1

Criteria library

Reusable inclusion and exclusion criteria sets — build once, apply to your next review.

F02.2

Automatic distribution

Papers split across reviewers with workload balancing — nobody hand-assigns 4,700 rows.

F02.3

Keyboard-first screening

Include, exclude, and advance without touching the mouse.

F02.4

AI suggestions

A decision, a 0–100 confidence score, and exactly which of your criteria drove it.

F02.5

AI tiers you set

Criteria-only from paper one, or guided by your own resolved decisions as worked examples.

F02.6

Full-text escalation

When the abstract isn’t enough, an optional AI re-analysis reads the entire PDF.

F02.7

Per-reviewer AI control

The owner decides, person by person, who screens with AI assistance. The decision stays yours.

F03 / RESOLVE CONFLICTS

Disagreements surface themselves.

5 CAPABILITIES

F03.1

Automatic detection

The moment two reviewers disagree, a conflict opens — nothing gets buried in a spreadsheet.

F03.2

Resolution workspace

Both votes, both sets of notes, and both reviewers’ PDF highlights in one view.

F03.3

AI-proposed resolutions

A reasoned proposal built from everything both reviewers saw — never an automatic decision.

F03.4

Full audit trail

Every vote, note, and final call is recorded — ready for the methods section.

F03.5

Configurable thresholds

You define what counts as a conflict for your project.

If it’s on this sheet, it’s in the product. Nothing here is a preview, a beta flag, or a promise.

CAPABILITY SHEET · CHECKED AGAINST THE CURRENT BUILD

F04 / EXTRACT

Your form, your fields — with receipts.

6 CAPABILITIES

5 FIELD TYPES

F04.1 — Custom templates

Text, number, dropdown, multi-select, and yes/no fields — shaped to your research question.

EVIDENCE

F04.2 — AI pre-fill

Every field it can fill comes with the exact sentence it came from, mapped to the source PDF.

PDF

F04.3 — Annotate while extracting

Highlight and comment on the PDF without leaving the form.

DIFF

F04.4 — Second-reviewer verification

An optional verification pass with a field-by-field diff view.

REUSE

F04.5 — Template library

Save extraction templates and reuse them across projects.

FLAG

F04.6 — Notes & flags

Leave a note or flag any response for review — small doubts don’t get lost.

F05 / REPORT

From decisions to a submission-ready draft.

5 CAPABILITIES

F05.1

Evidence table

One row per included study — sortable, filterable, CSV export.

F05.2

AI Methods draft

Drafted from your actual criteria, screening setup, and inter-rater agreement.

F05.3

AI Results draft

Synthesized from your extraction data, with numeric [n] citations that resolve to your included studies.

F05.4

PRISMA 2020 diagram

Auto-computed from your project’s real numbers, across all phases. DOCX export.

F05.5

DOCX export

The report leaves the tool the way journals want it.

F06 / COLLABORATE & CONTROL

The right access for every role.

5 CAPABILITIES

SCOPEOWNERASSISTANTREVIEWER
Project settings & criteria
AI access & credit budget
Screening & voting
Conflict resolution
Extraction & verification
Reports & export

FULL CONTROL ASSIGNED SCOPE○ READ-ONLY

F06.1

Three clear roles

Owner, assistant, reviewer — each sees exactly what they need, nothing more.

F06.2

Email invitations

Co-reviewers join from a single email — no account gymnastics.

F06.3

Per-reviewer AI toggle

Enable or disable AI assistance per person — useful for blinded passes.

F06.4

Notifications

In-app and email, for the moments that matter — assignments, conflicts, finished reports.

F06.5

Archive mode

Finished reviews go read-only — the record stays intact, nothing gets edited by accident.

F07 / SCALE & SECURITY

Built for real corpora.

5 CAPABILITIES

F07.1

Tested at 5,000+ papers. Multi-reviewer projects at real review scale — pages stay fast when the corpus doesn’t.

F07.2

Encrypted, end to end. Storage and transport, no exceptions.

F07.3

Permissions enforced on the server. Access rules hold even against direct API calls — not just hidden buttons.

F07.4

An AI usage dashboard. See exactly what AI work each project and each person has used — no black box, no surprise bill.

F07.5

Your data, always. Export or delete your data anytime. It’s your review, not ours.

SEVEN STAGES · ONE AUDIT TRAIL

Stop stitching tools together.

Import to PRISMA to a drafted report — one place, one record. The free plan covers a full solo pass.