Skip to main content
We now turn to the cutting edge: designing and building that can orchestrate complex, multi-step research processes. This is about architectural thinking. We’ll explore how to design research workflows where AI agents work together, each taking on specialized roles, coordinating their efforts, and producing sophisticated outputs. More importantly, we’ll develop critical thinking about when such systems are appropriate, how to maintain quality control, and where human judgment remains irreplaceable. By the end of this guide, you’ll move from AI user to AI architect.

​
Learning Outcomes

By the end of this session, you’ll be able to:
  • Design multi-agent workflows where different AI roles collaborate on complex research tasks
  • Implement quality control frameworks that maintain research rigor in automated systems
  • Critically evaluate when agentic approaches add value and when simpler methods work better
  • Navigate the ethical and professional considerations of increasingly autonomous research systems
  • Build practical, cost-effective workflows that integrate into your academic research practice

​
Multi-Agent Research Architecture

MULTI-AGENT RESEARCH ARCHITECTURE

👤 HUMAN ARCHITECT (Research Orchestrator)
    |
    +-- WORKFLOW DESIGN PHASE:
        |
        +-- 📋 Define Research Goal
        +-- 🔍 Identify Sub-Tasks
        +-- 🎭 Assign Agent Roles
        +-- ✅ Set Quality Checkpoints
            |
            v
    🤖 SPECIALIZED AI AGENT TEAM:

        Agent 1: 🔍 DISCOVERY
            - Literature search
            - Citation mapping
            - Tools: Zotero MCP, Web Search MCP
            - Quality Gate: ✓ Relevance Check

        Agent 2: 📊 ANALYSIS
            - Pattern recognition
            - Theme extraction
            - Tools: Filesystem MCP, Sequential Thinking MCP
            - Quality Gate: ✓ Logic Validation

        Agent 3: 🧠 SYNTHESIS
            - Framework building
            - Theory integration
            - Tools: Sequential Thinking MCP
            - Quality Gate: ✓ Novelty Assessment

        Agent 4: 🎓 CRITIQUE
            - Quality validation
            - Gap identification
            - Tools: Zotero MCP, Filesystem MCP
            - Quality Gate: ✓ Integration Review

            |
            v (All Quality Gates Pass)
    📄 RESEARCH OUTPUT (Validated Synthesis)
            |
            +-- Feedback & Refinement --> back to Human Architect

🔧 MCP SERVER TOOLS (Available to all agents):
    - 📚 Zotero MCP (Library access)
    - 🌐 Web Search MCP (Real-time discovery)
    - 📁 Filesystem MCP (File operations)
    - 🧮 Sequential Thinking MCP (Reasoning chains)
Using ASCII Diagrams: Copy this architecture into your AI chat sessions when designing agentic workflows! Perfect for explaining your multi-agent system design.
Critical Distinction: Agentic workflows are not about removing humans from research - they’re about strategic delegation of cognitive labor. The human architect maintains oversight, designs the system, and validates outputs. Quality control gates ensure rigor is never compromised for automation.

​
Key Concepts: The Workshop Prep Kit

The following concepts provide the conceptual blueprint for agentic systems, the practical manual for building them, and the professional context for why this work matters.
  • Conceptual Architecture (“Towards an AI co-scientist”): This provides the high-level vision, moving from single prompts to orchestrating a team of specialized AI roles. This is the “what we are trying to build” model.
  • Engineering Reality (“How we built our multi-agent research system”): This is the “what can go wrong” guide. Learning from the hard-won lessons of professional teams helps anticipate and solve common problems.
  • Practical Implementation (“Claude Code: Best practices”): This is the primary lab manual, providing specific, actionable instructions and code patterns for hands-on work.
  • Professional Context (“From Scarcity to Abundance”): This provides the strategic context, helping to understand how agentic workflows are reshaping the future of academic research.

​
Building Your Research Architecture: A Step-by-Step Guide

Part A: Failure Analysis

Review Failure Patterns: Before building a complex system, analyze the common failure modes documented in the Failure Museum.Identify Quality Control Checkpoints: These documented failures become the quality control checkpoints for your new, more robust system.
Start here - understanding what goes wrong is the foundation for building robust systems.

Part B: Quality & Ethics

Ethical Considerations: Review institutional policies on AI use, data privacy, and proper documentation of AI assistance.Build a Quality Control Framework: Define the roles for human and AI at each stage of the research process (e.g., Discovery, Curation, Analysis, Writing) and specify the verification method for each stage.
Don’t skip this - ethical frameworks protect both you and your research integrity.

Part C: Multi-MCP Orchestration

Build a Research Agent System: Use a tool like Claude Code to connect multiple MCP servers (e.g., filesystem, web search).Chain Servers for Workflows: Design automated workflows that pass information between different servers to solve problems identified in your failure analysis.
This is where theory meets practice - time to build your agentic system!
Hands-On Implementation: For a complete, step-by-step guide to building an SLR with Claude Code (including screening, extraction, and synthesis workflows), see Building an SLR with Claude Code.

Recommended Exercise: Design Your Own Agentic Workflow

  1. Define Your Use Case: Build upon a systematic review or another research task.
  2. Design Your Architecture: Map out which tasks will be automated vs. human-controlled.
  3. Specify Your Agents: Define the roles and prompts for each AI agent in your system.
  4. Plan Quality Controls: Identify verification points and failure modes, using your documented failures as a guide.
  5. Calculate Costs: Estimate API usage and time savings.
This is the capstone exercise for applying agentic thinking to your own research.

​
Pre-Class Setup for Session 4

Before this session, please ensure your research environment is fully prepared:
1

Curate Your Knowledge Base

Have all your curated papers loaded and processed in your Cherry Studio knowledge base.
2

Test Your Workflow

Verify that the conversation export functionality to Obsidian is working correctly.
3

Prepare Your Protocol

Bring your draft systematic review protocol. We will be using it as the foundation for building our agentic workflows.

​
Beyond the Guide

The workflows you design here can become the foundation for dissertation research, collaborative projects, and professional research practice. Key Takeaway: Agentic AI is not about replacing human expertise—it’s about amplifying it through thoughtful delegation and orchestration.

​
Go Deeper

Ready to explore advanced AI capabilities?

Agentic Workflows

Build multi-agent research systems

MCP Servers

Extend AI capabilities with tools
Navigation: Previous: Session 3 • Return to Case Study Overview • Return to Homepage
⌘I