Have a look at this excellent guide here.

A Systematic Literature Review (SLR) is a structured method of finding, evaluating, and analyzing existing research on a specific topic. Unlike a traditional literature review, it follows a predefined and transparent process to minimize bias and ensure that all relevant studies are considered. The goal is to summarize the current state of knowledge, identify research gaps, and provide evidence-based conclusions.

Been noticing that AI-assisted SLR papers are increasing in numbers these few years as AI provides great help in organizing and collection information quickly in a previously manual task. 

SLRs have great potential in getting lots of citations as many researchers look for and refer to SLRs for a comprehensive roadmap before embarking on their individual research directions.

Important to note that I'm not an expert. Experts from our Gelombang AI team specializing in this are Dr. Noor Fitrah Abu Bakar and Dr. Azliyana Azizan. Do invite them to give a talk in your organization for advanced techniques and methods.

The general workflow on SLRs are:

Research Question → Generate Question Set → Evidence Files → Evidence Matrix → AI Synthesis → SLR Draft → Human Revision 

Example:

Engineering: Machine Learning for Predictive Maintenance

Social Science: Social Media Influence on Political Participation

Humanities: Digital Preservation of Cultural Heritage

Art & Design: Generative AI in Creative Design Education

Use this prompt to prepare questions to search for:

You are an expert systematic literature review (SLR) methodology consultant. I am preparing a systematic literature review on the topic:

[TOPIC]

Research domain:
[ENGINEERING / SCIENCE / SOCIAL SCIENCE / HUMANITIES / ART & DESIGN]

Research objective:
[OBJECTIVE]

Your task is NOT to write the literature review yet. Instead, create a complete evidence collection plan.

Generate:

1. Core concepts and keywords.

2. Alternative keywords and synonyms.

3. Boolean search strings suitable for Scopus.

4. Inclusion criteria.

5. Exclusion criteria.

6. Publication year range recommendations.

7. Research questions suitable for an SLR.

8. A list of evidence extraction questions that I should ask Scopus AI.

The evidence extraction questions must be designed so that the answers can later be used to write:

• Introduction

• Problem statement

• Research gap

• Methodology

• Findings

• Thematic analysis

• Future research directions

• Conclusion

For each evidence extraction question, explain:

• Why it is important

• Which section of the SLR it supports

• Suggested filename to store the answer

Output the result as a structured evidence collection protocol. Use a standardized naming scheme:

• 01_Definitions.txt

• 02_Concepts.txt

• 03_Evolution.txt

• 04_Methodologies.txt

• 05_Theories.txt

• 06_Findings.txt

• 07_Challenges.txt

• 08_Gaps.txt

• 09_FutureDirections.txt

• 10_KeyAuthors.txt

• 11_InfluentialPapers.txt

• 12_ResearchAgenda.txt

Suppose the topic is Machine Learning for Predictive Maintenance

The AI may generate questions like:

• What are the most commonly accepted definitions of predictive maintenance in the literature?

• How has predictive maintenance research evolved from 2015-2025?

• What machine learning methodologies are most frequently used?

• What datasets are commonly used and what are their limitations?

• What challenges and barriers are consistently reported?

• What research gaps are repeatedly identified by authors?

• What future research directions are recommended?

Now use Scopus AI. For every question:

• Ask question

• Copy answer

• Save each answer in different files. 

This becomes your evidence repository.

After collecting all files, upload all files into ChatGPT. Then use this prompt:

You are acting as a senior researcher and systematic literature review specialist.

I have uploaded evidence files collected from Scopus AI.

Your task is to synthesize the information into an academically rigorous systematic literature review. Requirements:

• Do not simply summarize each file.

• Identify recurring themes.

• Identify contradictions.

• Identify emerging trends.

• Identify theoretical and methodological gaps.

• Build a thematic framework.

• Compare findings across studies.

• Produce a coherent narrative.

Generate:

• Abstract

• Introduction

• Review Method

• Thematic Findings

• Research Gaps

• Future Research Directions

• Practical Implications

• Theoretical Implications

• Conclusion

For every claim, indicate which uploaded evidence file supports it.

Ask the AI:

Using the uploaded evidence files, create a literature review evidence matrix.

Columns:

• Theme

• Key Findings

• Supporting Evidence File

• Contradictory Evidence

• Research Gap

• Future Direction

Output as a table.

This matrix becomes the backbone of the Results and Discussion section.

After you have collected all the Scopus AI outputs and uploaded them to ChatGPT, use a prompt like the one below. It forces the AI to write the paper iteratively, section by section, instead of generating a weak one-shot review.

Prompt:

You are a senior academic researcher, journal reviewer, and systematic literature review (SLR) expert. I have uploaded evidence files extracted from Scopus AI. These files contain synthesized information from the literature and serve as the evidence base for this review. Your task is to help me write a complete publishable SLR paper. Rules:

1. Write ONE section at a time.

2. Do not proceed to the next section until I approve the current section.

3. Every section must be academically rigorous, coherent, and connected to previous sections.

4. Use only information supported by the uploaded evidence files.

5. Identify patterns, trends, agreements, contradictions, and research gaps across the evidence.

6. Do not simply summarize files independently; synthesize information across sources.

7. Maintain a formal academic writing style suitable for Scopus-indexed journals.

8. When making claims, indicate which evidence file(s) support the claim.

9. Highlight areas where evidence is insufficient or contradictory.

10. Suggest figures, conceptual frameworks, tables, and thematic maps whenever appropriate.

Follow this workflow:

STEP 1 – Analyze Evidence Base

* Review all uploaded files.

* Create a summary table showing:

  * Main themes

  * Key findings

  * Frequently cited concepts

  * Research gaps

  * Emerging trends

* Wait for approval.

STEP 2 – Develop Paper Structure

* Propose:

  * Title

  * Research objectives

  * Research questions

  * Major themes

  * Complete paper outline

* Wait for approval.

STEP 3 – Write Introduction

Include:

* Background

* Importance of the topic

* Current state of research

* Research problem

* Research gap

* Objectives

* Contributions of the review

* End with paper organization

* Wait for approval.

STEP 4 – Write Methodology

Include:

* SLR approach

* Search strategy

* Inclusion and exclusion criteria

* Data sources

* Screening process

* Data extraction process

* Data synthesis process

* PRISMA-style explanation

* Wait for approval.

STEP 5 – Write Findings and Thematic Analysis

For each theme:

* Define the theme

* Synthesize findings

* Compare studies

* Discuss agreements and disagreements

* Identify strengths and weaknesses

* Present supporting evidence

* End with a mini-summary

* Wait for approval after each theme.

STEP 6 – Write Discussion

Include:

* Cross-theme analysis

* Theoretical implications

* Practical implications

* Methodological implications

* Comparison with previous reviews

* Major insights

* Wait for approval.

STEP 7 – Write Research Gaps and Future Directions

Include:

* Confirmed gaps

* Emerging opportunities

* Proposed future research agenda

* Recommendations for researchers

* Wait for approval.

STEP 8 – Write Conclusion

Include:

* Summary of findings

* Contributions

* Limitations

* Final recommendations

* Wait for approval.

STEP 9 – Final Refinement

Generate:

* Abstract

* Keywords

* Graphical abstract description

* Highlights

* Contribution statement

* Suggested tables

* Suggested figures

* Journal-ready final manuscript structure

Begin with STEP 1 only. Do not write any other section until approval is given.

This prompt works well because it mirrors how experienced researchers actually write SLRs: evidence analysis → outline → introduction → methodology → themes → discussion → gaps → conclusion → abstract, rather than trying to generate the entire paper in a single pass. That staged approach usually produces a much stronger manuscript for researchers from different disciplines.

All the best!