Skip to content

>>> 003 / AI features

AI that shows its work.

Every suggestion Littview’s AI makes comes with a reason and a source. It never decides for you — it hands you a well-argued starting point.

P1

You stay in charge

AI output is always a suggestion with reasoning attached, never a silent decision. Accept, edit, or reject — every time.

P2

Evidence-linked

Suggestions cite the criterion, the passage, or the data point they came from — click through and see it in the source PDF.

P3

Transparent usage

A credits system and usage dashboard show exactly what AI work cost, per project and per person. No surprise bills.

P4

Per-reviewer control

The project owner decides who has AI assistance turned on, person by person.

003.1

AI Screening

Littview reads each paper's title and abstract against your project's actual inclusion/exclusion criteria and returns a decision, a 0–100 confidence score, and exactly which criteria drove it. It gets more tailored to your project as it sees more of your own resolved decisions.

  • Works from paper one — no cold start
  • Learns from your resolved decisions
  • Confidence + criteria on every suggestion
  • Optional full-text re-analysis
  • Never auto-decides
003.2

AI Conflict Resolution

When two reviewers disagree, Littview doesn't just flag it — its AI weighs both votes, both sets of notes, and both reviewers' PDF highlights, then proposes a resolution with its reasoning laid out. You see everything it saw.

  • Full context in one workspace
  • Reasoned resolution proposal
  • Complete audit trail
  • Human sign-off required
003.3

AI Data Extraction

Point Littview at your extraction form and it pre-fills every field it can from the source PDF — and shows you the exact sentence it used for each one. Nothing is a black box: click any field, see the evidence.

  • Field-by-field pre-fill
  • Evidence-mapped to source passage
  • Your custom form, not a fixed template
  • Optional second-reviewer verification
003.4

AI Report Writer

When your review is done, Littview drafts PRISMA 2020-compliant Methods and Results sections directly from your project's real data — your criteria, your screening setup, your inter-rater agreement, your extraction findings. Citations are numbered and resolve to your actual included studies.

  • Methods from your real configuration
  • Results synthesized from extraction data
  • Numeric citations + references list
  • DOCX export
  • Always a draft to edit, never a final answer

One credits system.
No surprise bills.

AI features run on a simple credits system — no separate bill per feature, no black box. Your usage dashboard shows exactly what each project and person has used.

See how credits work →
Q&A
Straight answers.
Does AI ever make the final decision?
No. Every AI output — a screening suggestion, an extracted field, a drafted paragraph — is something a human reviewer accepts, edits, or rejects. Nothing is applied automatically.
Can I turn AI off for specific reviewers?
Yes. The project owner can enable or disable AI assistance per person.
Is our data used to train AI models?
No — Littview's AI runs on OpenAI's API, which does not use API data to train its models.
What happens if AI gets it wrong?
You review and correct it like any other suggestion — every AI output is fully editable, and your correction is what's recorded as the final decision.

See it on your own references.

Start free
No credit card required