← TrendWatcher
arXiv cs.AI
6/10

ScholarLoop

An AI literature-review assistant for academic researchers that learns from your accept/reject decisions week after week, progressively narrowing its search strategy and citation screening to match your specific project's focus — no prompt engineering needed.

Target user

PhD students and postdocs running systematic literature reviews or building a literature chapter

Features
  • Auto-narrowing search that learns the keywords, journals and authors you keep accepting
  • Adaptive citation screening that mirrors your inclusion pattern as the project matures
  • PRISMA-ready export with flow diagram and screening log for your methods section
  • Disagreement digest surfacing papers where you and the AI diverge, so nothing slips through silently
Why now

Literature-review AI tools have exploded but every researcher re-pastes context every session; a project-specific co-evolving assistant is the missing layer.

Signals · overall 6/10
Demand
8/10

Multiple recent (2025-2026) head-to-head comparison articles ranking 5-17 AI literature-review tools, and Rayyan alone reports 1M+ researchers at 20,000+ institutions — clear, active demand.Best AI Tools for Systematic Literature Reviews in 2026: Compared for PhD StudentsRayyan Pricing

Whitespace
7/10

The persistent, accept/reject-driven learning loop is not explicitly offered by Elicit, Rayyan, Covidence, Sciwand, or ResearchRabbit — comparison guides highlight fragmented stacks and per-session re-pasting of context as an unsolved pain.Best AI Tools for Systematic Literature Reviews in 2026: Compared for PhD StudentsCovidence vs Rayyan vs Elicit vs INRA

Monetization
6/10

Proven willingness to pay at ~$10–12/mo for Elicit and ~$10/mo for Rayyan Pro, with $169/mo Elicit team tiers and institutional Covidence plans — viable SaaS pricing exists, but PhDs are price-sensitive and many competitors have free tiers.Pricing | ElicitRayyan Pricing

Longevity
7/10

Systematic literature reviews are a durable academic workflow tied to publishing norms (PRISMA), and AI-augmented research is a multi-year trend; however, the specific co-evolving memory moat is replicable by incumbents adding a feedback layer.Best AI Tools for Systematic Literature Reviews in 2026: Compared for PhD Students

Feasibility
4/10

Non-trivial: requires multi-database ingestion (PubMed, Scopus, OpenAlex, Semantic Scholar), PRISMA workflow support, PDF parsing, Zotero/EndNote interop, and a persistent per-project memory store — incumbents have years of integration work already shipped.Best AI Tools for Systematic Literature Reviews in 2026: Compared for PhD Students

Co-Harness: Co-Evolving Harnesses and Model Weights for LLM AgentsarXiv cs.AI · 2026-07-28 (today)