ChatGPT agent compiles an annotated bibliography
Sultan M. Quresh used ChatGPT agent to move from a broad academic topic to a structured set of sources and a working paper draft. He asked it to find recent peer-reviewed research about how big data and analytics inform decision-making. The agent searched academic sources, including PubMed, ScienceDirect, ResearchGate, Springer and PubMed Central, then filtered for relevance and the requested publication window.
Original prompt
Shared by the source author. See source notes for context.
Find recent peer-reviewed papers (from the past five years) that explore how Big Data and Analytics inform decision-making
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source poster / publisher
- Sultan M. Quresh
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- Quresh says a half-day research and drafting task took under ten minutes. He still recommends human verification of all sources and claims.
- Human involvement
- Quresh retained responsibility for verifying that papers existed, citations were accurate and interpretations matched the originals.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Quresh defined a topic and limited the search to recent peer-reviewed studies.
- The agent searched research sources and compiled APA-style annotations with methods, findings and relevance.
- It synthesized themes and gaps into draft material, which Quresh says he checked against the original papers.
Sultan M. Quresh used ChatGPT agent to move from a broad academic topic to a structured set of sources and a working paper draft. He asked it to find recent peer-reviewed research about how big data and analytics inform decision-making. The agent searched academic sources, including PubMed, ScienceDirect, ResearchGate, Springer and PubMed Central, then filtered for relevance and the requested publication window.
For each selected paper, Quresh says the agent prepared an annotated bibliography with an APA-style citation, a note about the method and findings, and an explanation of relevance. It also grouped themes and research gaps into a synthesis, then produced material that could be exported from Markdown into a Word document. Quresh reports that a task he expected to take half a day took under ten minutes.
That speed was a first-draft result, not proof that every reference or interpretation was correct. Quresh explicitly recommends checking source links, citations and claims, and reviewing the synthesis before using it in academic work. The transferable workflow is to give the agent a narrow topic and date range, ask for traceable annotations, and verify each paper against its original publication before relying on the output.
Prompt
Original prompt
Source: Medium · Sultan M. Quresh
Frequently asked questions
What was the research topic?
Quresh asked for recent peer-reviewed papers on how big data and analytics inform decision-making.
What did the agent produce?
It prepared annotated citations, summaries of methods and findings, a thematic synthesis and draft material that could be exported to Word.
Did Quresh verify the research output?
He recommends checking the source links, citations and interpretations. The reported time saving does not remove that review step.
More details
About this account
The agent searched sources, summarized papers and assembled draft research material from Quresh’s topic and constraints.
This is an exact excerpt from the author’s published prompt, shortened to 17 words.
Credit: Sultan M. Quresh — Medium
Source published
Added to Agents In Use
Source checked
Tools and topics
ChatGPT agent · PubMed · ScienceDirect · ResearchGate · Springer · PubMed Central · Markdown · Microsoft Word
academic research · annotated bibliography · literature review
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.