r8x-ai Research Engine

Agentic AI-Powered Research Analysis Platform

Transform research papers into structured insights, identify knowledge gaps, generate novel hypotheses, and build interactive knowledge graphs using a multi-agent AI workflow.

---

🎥 Demo

Full Working Video Link

https://drive.google.com/file/d/1rtGNIa4nCM5GOxXunKJNEkJA8toOPZ16/view?usp=sharing

📖 Overview

R8X-AI Research Engine is an advanced Agentic AI platform that automates academic research workflows using a collaborative multi-agent architecture.

Unlike traditional AI chat applications that rely on a single LLM prompt, R8X-AI decomposes research into specialized tasks:

• Research Planning
• Paper Ingestion
• Information Extraction
• Gap Analysis
• Hypothesis Generation
• Novelty Validation
• Knowledge Graph Construction
• Automated Report Generation

The system combines multiple AI agents, external research databases, vector search, and structured storage to deliver explainable and traceable research outputs.

---

✨ Features

🤖 Multi-Agent Research Workflow

11 specialized AI agents collaborate to perform research tasks.

📚 Research Paper Analysis

• Semantic Scholar Integration
• arXiv Integration
• PDF Upload Support
• Automatic Metadata Extraction

🧠 Knowledge Extraction

• Claim Extraction
• Entity Recognition
• Methodology Detection
• Citation Mapping

🔍 Research Gap Discovery

Identify:

• Contradictions
• Missing Connections
• Underexplored Areas
• Future Research Opportunities

💡 Hypothesis Generation

Generate novel research ideas based on discovered knowledge gaps.

🛡️ Hypothesis Critique

Validate generated ideas against prior research to reduce duplication and hallucinations.

🌐 Interactive Knowledge Graph

Visualize relationships between:

• Papers
• Concepts
• Claims
• Methods
• Hypotheses

📄 Automated Report Generation

Generate:

• Markdown Reports
• Research Summaries
• Structured Findings
• LaTeX Documents

---

🏗️ Architecture

User Query
    │
    ▼
Guardrail Agent
    │
    ▼
Planner Agent
    │
    ▼
Executor Agent
    │
 ┌──┴─────────────────────────┐
 │                            │
 ▼                            ▼
Ingestion Agent         PDF Reader Agent
 │
 ▼
Structuring Agent
 │
 ▼
Gap Analysis Agent
 │
 ▼
Hypothesis Agent
 │
 ▼
Critic Agent
 │
 ▼
Graph Agent
 │
 ▼
Report Agent
 │
 ▼
Final Research Report

---

🤖 AI Agent Ecosystem

Agent	Responsibility
Guardrail Agent	Detects research intent
Planner Agent	Creates execution plan
Executor Agent	Orchestrates workflow
PDF Reader Agent	Extracts document content
Ingestion Agent	Fetches research papers
Structuring Agent	Extracts claims & entities
Gap Analysis Agent	Finds research gaps
Hypothesis Agent	Generates novel ideas
Critic Agent	Validates hypotheses
Graph Agent	Creates knowledge graph
Report Agent	Generates final report

---

🛠 Tech Stack

Frontend

• Next.js 16
• React 19
• TypeScript
• Tailwind CSS 4
• Radix UI
• Vis Network

Backend

• Next.js API Routes
• Node.js

AI Layer

• OpenAI Agents SDK
• Vercel AI SDK
• LangChain
• OpenAI GPT Models
• Anthropic Claude

Database

• Supabase PostgreSQL
• pgvector

Storage

• Cloudflare R2
• Upstash Redis

Data Processing

• Puppeteer
• PDF Parse

---

🔄 Research Pipeline

1. Research Planning

Breaks complex research goals into actionable tasks.

2. Academic Data Collection

Retrieves papers from multiple research sources.

3. Knowledge Structuring

Extracts:

• Claims
• Findings
• Entities
• Methodologies

4. Gap Analysis

Identifies:

• Contradictions
• Weaknesses
• Missing Research

5. Hypothesis Creation

Generates novel research directions.

6. Validation

Checks novelty against existing literature.

7. Knowledge Graph Construction

Creates relationship graphs.

8. Report Generation

Produces comprehensive research documentation.

---

📂 Project Structure

r8x-ai-research-engine
│
├── app/
│   ├── api/
│   └── dashboard/
│
├── components/
│
├── lib/
│   ├── agents/
│   │   ├── guardrail.ts
│   │   ├── planner.ts
│   │   ├── executor.ts
│   │   ├── ingestion.ts
│   │   ├── pdf-reader.ts
│   │   ├── structuring.ts
│   │   ├── gap-analysis.ts
│   │   ├── hypothesis.ts
│   │   ├── critic.ts
│   │   ├── graph-agent.ts
│   │   └── report.ts
│   │
│   └── db/
│
├── public/
│
└── supabase/

---

⚡ Installation

git clone https://github.com/Nihal-kappungal/r8x-ai-research-Engine.git

cd r8x-ai-research-engine

pnpm install

Create environment variables:

OPENAI_API_KEY=
ANTHROPIC_API_KEY=

SUPABASE_URL=
SUPABASE_ANON_KEY=

UPSTASH_REDIS_REST_URL=
UPSTASH_REDIS_REST_TOKEN=

R2_ACCESS_KEY_ID=
R2_SECRET_ACCESS_KEY=
R2_BUCKET_NAME=

Run development server:

pnpm  dev

---

📊 Example Research Query

Analyze recent advancements in solid-state batteries,
identify charging cycle limitations,
and propose a novel methodology to overcome them.

Generated Outputs

✅ Research Papers

✅ Extracted Claims

✅ Research Gaps

✅ Novel Hypotheses

✅ Knowledge Graph

✅ Final Research Report

---

🎯 Key Differentiators

Unlike traditional AI chatbots:

• Multi-Agent Architecture
• Research-Oriented Workflow
• Real Academic Data Sources
• Structured Knowledge Extraction
• Hypothesis Generation
• Knowledge Graph Visualization
• Traceable Research Process

---

🚀 Future Roadmap

• Multi-modal research support
• Research collaboration workspace
• Citation generation
• Local LLM support
• Agent memory system
• Research benchmarking
• Automated literature reviews

---

📜 License

MIT License

---

👨‍💻 Author

Nihal K

Final Year B.Tech CSE Student

Built as a research-focused AI system for automating academic analysis through multi-agent collaboration.

---

⭐ If you found this project useful, consider giving it a star!