Transform your research workflow with AI orchestration
From manual literature reviews → Automated, systematic, scalable research systems
A complete system for academic researchers and scientists to leverage AI for:
- 📚 Automated Literature Reviews - Process 100+ papers in days, not months
- 🎯 Systematic Evaluation - Custom rubrics for consistent quality assessment
- 🧠 Intelligent Synthesis - AI helps identify patterns, gaps, and insights
- ⚡ Research Automation - Scripts and workflows for repetitive tasks
- 🔧 AI Research Workbench - Coordinated toolset for modern research
This pathway starts from the same foundation (USER-PROFILE.md) but focuses entirely on research, not software development.
- Complete main USER-PROFILE.md
- Add research-specific sections: RESEARCH-PROFILE-ADDITIONS.md
- Save as
MY-RESEARCH-PROFILE.md
Follow AI Workbench Setup Guide:
- Install search tools (Scite.ai, Elicit, Consensus)
- Set up analysis AI (Claude Chat, ChatGPT)
- Configure citation management (Zotero)
- Organize note-taking system
Use Prompt R1: Research Roadmap Generator:
- Input your research question and goals
- Get personalized structured learning path
- Tool recommendations for your domain
- Milestone projects to validate progress
Use Prompt R2: Literature Evaluation Framework:
- Generate custom rubrics for your field
- Create screening criteria
- Build extraction templates
- Establish quality standards
Use Prompt R3: Research Orchestration:
- Coordinate multiple AI tools
- Run systematic literature reviews
- Automate paper discovery and evaluation
- Synthesize findings at scale
- 🔬 RESEARCH-PROFILE-ADDITIONS.md - Research-specific profile template
- 🎯 PROMPTS/ - Three research prompts:
- 🔧 TOOLS/ - Research AI Tool Matrix (Scite, Elicit, etc.)
- 📋 WORKFLOWS/ - AI Workbench Setup, systematic review workflows
- ⚡ QUICK-REFERENCE/ - Research quick guides
- 🌱 DOMAIN-ADAPTATIONS/ - Field-specific guides
Traditional Research (Inefficient):
You → Google Scholar → Read 100 papers manually → Takes months
↓
Overwhelmed, miss papers, inconsistent evaluation
Orchestrated Research (Optimal):
You (Conductor) → Scite.ai (Citation analysis)
→ Elicit (Semantic search)
→ Perplexity (Current papers)
→ Claude Chat (Deep analysis)
→ Scripts (Automation)
↓
Process 100+ papers systematically in weeks
- Law of Specialization - Use Scite for citations, Claude for synthesis, Perplexity for current info
- Law of Context Transfer - You carry research context between AIs (your profile, rubric, questions)
- Law of Human Judgment - AI handles volume, you provide expertise and critical evaluation
When: Starting new research area, need systematic approach AI: Perplexity or Claude Chat Output: Structured learning path for AI-augmented research
When: Need consistent paper assessment criteria AI: Claude Chat or ChatGPT Output: Custom rubrics, extraction templates, quality standards
When: Running literature reviews, managing high paper volume AI: Multi-tool coordination (Scite + Claude + Perplexity + Scripts) Output: Complete research pipeline from discovery → synthesis
- Literature review for dissertation
- Staying current in fast-moving fields
- Identifying research gaps systematically
- Grant proposal literature sections
- Systematic reviews and meta-analyses
- Paper writing with strong lit reviews
- Grant writing with comprehensive backgrounds
- Peer review with systematic evaluation
- Monitoring emerging research trends
- Interdisciplinary literature synthesis
- Large-scale evidence reviews
- Team research coordination
- Competitive intelligence
- Technology landscape analysis
- Evidence-based decision making
- Patent and publication monitoring
| Tool Type | Recommended Tools | Purpose |
|---|---|---|
| Search & Discovery | Scite.ai, Elicit, Consensus, Google Scholar | Finding relevant papers |
| Analysis & Synthesis | Claude Chat, ChatGPT, Perplexity | Deep reading, synthesis, writing |
| Citation Management | Zotero, Mendeley | Organizing references |
| Knowledge Management | Notion, Obsidian, Spreadsheets | Notes and synthesis |
| Visualization | Research Rabbit, Connected Papers | Citation networks |
| Automation | Python scripts, Zapier, n8n | Monitoring, extraction |
See full comparison: Research AI Tool Matrix
- Uses Google Scholar only
- Reads papers one by one
- Takes scattered notes
- 10-20 papers = 1-2 weeks
- Uses 2-3 search tools
- AI helps with summaries
- Basic organization
- 20-30 papers = 1 week
TARGET LEVEL - Most researchers stabilize here
- Uses 5+ tools fluently
- Systematic evaluation with rubrics
- AI-human collaboration smooth
- 50-100 papers = 1-2 weeks
- Full orchestration workflow
- Scripts for monitoring/extraction
- Teaching others the system
- 100+ papers = 1-2 weeks
Before AI Orchestration:
- Literature review for grant: 3 months
- Read 80 papers manually
- Missed key citations
- Synthesis felt overwhelming
After (Week 8 of this system):
- Same type of review: 3 weeks
- Processed 150 papers systematically
- Found papers Google Scholar missed
- Clear thematic synthesis
- Result: Grant funded, extra time for actual research
See complete research journeys →
RESEARCH-PATHWAY/
├── README.md ← You are here
├── RESEARCH-PROFILE-ADDITIONS.md ← Add to main profile
│
├── PROMPTS/ ← The Research Trinity
│ ├── prompt-r1-research-roadmap.md
│ ├── prompt-r2-literature-evaluation.md
│ └── prompt-r3-research-orchestration.md
│
├── WORKFLOWS/ ← Step-by-step guides
│ ├── ai-workbench-setup.md ← Start here for setup
│ ├── systematic-review-workflow.md
│ ├── meta-analysis-workflow.md
│ └── paper-writing-workflow.md
│
├── TOOLS/ ← Resources
│ ├── research-ai-tool-matrix.csv
│ ├── research-scripts/ ← Python examples
│ └── evaluation-rubric-templates/
│
├── QUICK-REFERENCE/ ← Print these!
│ ├── research-tool-selector.md
│ └── quality-checklist.md
│
└── DOMAIN-ADAPTATIONS/ ← Field-specific
├── life-sciences.md
├── social-sciences.md
├── physical-sciences.md
└── humanities.md
Research-specific contributions welcome:
- 📚 Domain Adaptations - Evaluation criteria for your field
- 🔬 Workflow Patterns - Your research orchestration patterns
- 📊 Success Stories - Share results and time savings
- 🔧 Tool Updates - New research AI tools and integrations
- 📜 Scripts - Automation examples for common tasks
See CONTRIBUTING.md for details.
Always:
- ✅ Verify AI-generated citations (check they exist!)
- ✅ Deep-read key papers yourself
- ✅ Use AI for efficiency, not to skip thinking
- ✅ Be transparent about AI use in methods sections
- ✅ Maintain research rigor and critical evaluation
Never:
- ❌ Trust AI citations without verification
- ❌ Skip reading papers entirely
- ❌ Let AI make final judgments on quality
- ❌ Use AI-generated text without thorough editing
- ❌ Compromise research integrity for speed
This system helps you:
- Process MORE papers MORE thoroughly
- Find papers you'd otherwise miss
- Apply consistent evaluation criteria
- Identify patterns across large literature sets
- Free up time for deep thinking and analysis
Goal: Better research, faster. Never faster research, lower quality.
License: MIT - Share freely, adapt extensively
Created by: Magnus Smari Smarason Version: 1.0 - Research Pathway Year: 2025
Part of: Taming Technology - AI Learning Orchestration System
- Complete RESEARCH-PROFILE-ADDITIONS.md
- Set up 2-3 tools from Workbench Setup
- Use Prompt R1 to get roadmap
- Start with 10 papers to test workflow
- Read complete guide
- Set up full AI Research Workbench
- Create evaluation rubric
- Follow complete orchestration workflow
- Development Pathway - For software development (not research)
- Main Guide - Full AI orchestration philosophy
- User Profile - Foundation for all pathways
- Research questions? Open a discussion
- Found an issue? Report it
- Success story? Share it!
"You're not just learning to use AI tools. You're learning to orchestrate intelligence—both artificial and your own—to advance knowledge that was previously out of reach."
You have an entire orchestra of research AI tools ready to help. You're the conductor. This is your research symphony.
Now go advance knowledge. 🔬📚🧠📊🎓
⭐ Don't forget to star this repo if you found it helpful!