Run a systematic literature review from one prompt
Systematic literature reviews take 12-18 months to complete. Task: Suggested implementation: Use SciSpace's autonomous agent to initiate a systematic literature review via a single prompt. The agent generates PICO-based research questions and screening criteria, which the user must review and edit. After the user selects databases and date ranges, the agent automates searching, reranking, title/abstract screening, full-text downloading, data extraction, and final report generation including a PRISMA diagram.; tools: custom-built or unnamed agent. Human role: The original post does not say how a person stayed involved.; limitations: Prerequisites include access to SciSpace and relevant academic databases. A major constraint is that papers behind paywalls cannot be fully automated, limiting completeness. The source is a paid partnership, which may bias claims of utility. Human approval is required for research questions, screening criteria, and database selection before automation begins..
Tools
custom-built or unnamed agent
Published on Agent Work Atlas:
Overview
Systematic literature reviews take 12-18 months to complete.
Workflow
Suggested implementation: Use SciSpace's autonomous agent to initiate a systematic literature review via a single prompt. The agent generates PICO-based research questions and screening criteria, which the user must review and edit. After the user selects databases and date ranges, the agent automates searching, reranking, title/abstract screening, full-text downloading, data extraction, and final report generation including a PRISMA diagram.
Reported result
The source does not provide measured outcomes, accuracy rates, or time savings for this specific workflow. It only claims the process is automated after initial human input. A measurable acceptance check would be comparing the agent's PRISMA diagram and included paper list against a manual review of the same topic to verify completeness and accuracy.
Human role
The original post does not say how a person stayed involved.
Limitations
Prerequisites include access to SciSpace and relevant academic databases. A major constraint is that papers behind paywalls cannot be fully automated, limiting completeness. The source is a paid partnership, which may bias claims of utility. Human approval is required for research questions, screening criteria, and database selection before automation begins.
Try it
Act as a research assistant using SciSpace. I want to conduct a systematic literature review on [replaceable topic, e.g., 'AI in healthcare diagnostics']. First, generate PICO-based research questions and screening criteria for my review. Do not proceed until I approve these. Once approved, ask me to select databases and date ranges. Then, execute the automated search, screening, and data extraction. Finally, produce a PRISMA diagram and a table of included papers with rationales. Before finalizing, show me a sample of 5 screened papers for my verification. If paywalled papers are missing, list them separately. Output the final review in [replaceable format, e.g., Markdown].
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.