AI Essentials Course — Phase 4: Mastery and Application
Session 16: Building a Research Workflow with AI
Combine Perplexity, Claude, and ChatGPT into an integrated research workflow — from source discovery to annotated bibliography — that will serve you throughout your academic career.
Learning Objectives
What You'll Learn
By the end of this session, you will be able to:
- Design a multi-platform AI research workflow combining Perplexity, Claude, and ChatGPT
- Use Perplexity to discover and gather sources on a research question
- Use a writing AI to create an annotated bibliography from your gathered sources
- Understand how to chain the outputs of one AI into the inputs of another
Platform Access
Getting Started with Perplexity
Follow these steps to access Perplexity and get ready for today's lesson.
- Go to https://perplexity.ai and sign in.
- Also have ChatGPT and/or Claude open in separate tabs — you'll use multiple platforms in sequence today.
- Choose a specific research question you want to explore — something concrete enough that you could imagine writing a paper about it. For example: 'How do community-based programs support social engagement in adults over 70?' or any question in your academic field.
- Have a document open to capture notes, source information, and the outputs you'll generate in each step.
- Today's session has more steps than most — take your time and follow the workflow sequence. Each step builds on the previous one.
Free Account Required
All platforms used in this course offer free accounts with no credit card required. If you already have an account, simply sign in. The free tier gives you everything you need to complete this session.
Core Lesson
Today's Lesson
Read through this lesson carefully before starting the practice exercises below.
You've now spent fifteen sessions learning individual AI skills — structuring prompts, summarizing documents, researching with citations, editing writing, analyzing literature. Today you put it all together. This session is about workflow: how to combine the different tools you've learned into a coordinated system that makes your academic research more efficient and rigorous.
A multi-AI research workflow has three main stages, each powered by the tool best suited for it. Stage 1: Discovery — Use Perplexity to find sources, because it searches the web and cites what it finds. Stage 2: Analysis — Use Claude to analyze and synthesize the sources you've gathered, because it handles complex, nuanced analytical tasks exceptionally well. Stage 3: Writing — Use ChatGPT to help organize and draft the writing artifacts you need, like annotated bibliographies, outlines, or summaries. Each tool does what it does best, and their outputs chain together naturally.
Think of this workflow like a research team with specialized roles. Perplexity is the research librarian — fast, knowledgeable about what's available, and helpful for pointing you to the right sources. Claude is the research analyst — careful, nuanced, able to synthesize and critique complex material. ChatGPT is the writing collaborator — generative, flexible, good at producing and organizing text. Together, they're more powerful than any one of them alone.
The key insight in multi-platform chaining is that the output from one tool becomes the input to the next. You take what Perplexity found, paste the key information into Claude for analysis, and then take Claude's synthesis into ChatGPT for writing. Each handoff adds a layer of processing and refinement. This is the kind of sophisticated AI use that separates novices from experts.
An annotated bibliography is one of the most useful academic artifacts you can create with this workflow. In an annotated bibliography, each source is followed by a short paragraph — typically 100–200 words — that summarizes what the source says and explains its relevance to your research. This document becomes the raw material for your literature review chapter and proves to your committee that you've engaged seriously with your sources.
As you build this workflow today, pay attention to where each handoff happens and why. Understanding the logic of which tool to use at which stage is the deeper skill — and it's one you'll apply every time you start a new research project.
Hands-On Practice
Practice Exercise
Follow these steps in Perplexity. Take your time — there's no rush. Learning happens through doing.
- In Perplexity, type your research question: "What does recent academic research say about [your research question]? Please provide at least 5 credible sources with citations." Record the sources Perplexity provides.
- Verify at least 2 of the sources by clicking through to the original links. Confirm they exist and match what Perplexity described.
- Open Claude and paste the source information. Ask: "Here are summaries of 5 sources on [your topic]. What are the main themes across these sources, and where do the authors agree and disagree?"
- Open ChatGPT and ask: "Please help me create an annotated bibliography in APA format for the following 5 sources. For each source, write a 100-word annotation that summarizes the key argument and explains its relevance to my research on [your topic]: [paste source info]"
- Review the annotated bibliography. Revise any annotations that seem inaccurate or that miss the key argument of the source.
- Reflect on the workflow: Which stage felt most valuable? Where did you need to intervene with your own judgment? How would you improve this workflow for your next research project?
Try These
Example Prompts to Try
Copy any of these prompts directly into Perplexity and see what happens. Feel free to modify them to match your own academic interests.
Summary
Key Takeaways
- A multi-platform AI research workflow assigns each tool its area of strength: Perplexity for source discovery, Claude for analysis and synthesis, ChatGPT for writing tasks.
- Chaining AI outputs — using the output of one tool as the input to the next — creates a layered research process more powerful than any single tool.
- Annotated bibliographies are an excellent artifact to produce through this workflow, providing the raw material for literature review chapters.
- Understanding which tool to use at which stage — not just how to use each tool individually — is the mark of AI research fluency.
Multi-Platform Workflow Prompts — Chaining Outputs Across Tools
You've designed and executed a three-stage AI research workflow: discover with Perplexity, analyze with Claude, write with ChatGPT. This output-chaining approach — where each AI's result feeds the next stage — is the professional-level use of AI tools. This is how academics and researchers are using AI to dramatically accelerate their scholarly work.