README.md

metadata

title: ReallyDeepResearch
emoji: 🔎
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 5.44.1
app_file: app.py
pinned: false

ReallyDeepResearch

A multi-agent research system that conducts comprehensive topic analysis through parallel section exploration and intelligent fact synthesis.

What It Does

• Parallel Research Pipeline: 7 specialized agents per research section working simultaneously
• Self-Healing Research: Quality assessment with automatic iteration loops for incomplete findings
• Framework-Driven Analysis: Structured exploration via big-idea (market landscape) or specific-idea (business viability) frameworks
• Fact Deduplication: Cross-section intelligence with conflict detection and resolution
• Publishable Reports: Executive summaries with source verification and downloadable formats

How It's Different

Feature	ReallyDeepResearch	Perplexity Pro	OpenAI Deep Research
Research Depth	7-step pipeline per section	Single-pass queries	Chain-of-thought reasoning
Parallel Processing	Multi-section simultaneous	Sequential search	Linear processing
Quality Control	Critic agent + self-healing	Manual refinement	Built-in verification
Structured Output	Framework-specific analysis	General summaries	Problem-specific responses
Fact Verification	Cross-source validation	Source citations	Integrated fact-checking

Quick Start

Prerequisites

python 3.9+

Installation

git clone https://github.com/yourusername/ReallyDeepResearch
cd ReallyDeepResearch
pip install -r requirements.txt

Environment Setup

cp .env.sample .env
# Add your API keys:
# - OPENAI_API_KEY=your_key
# - SERPER_API_KEY=your_key  
# - DEFAULT_MODEL_NAME=gpt-4

Run Locally

python app.py

Open http://localhost:7860 → Enter topic → Choose framework → Get comprehensive research report.

Deploy to Hugging Face Spaces (Free)

See DEPLOYMENT.md for detailed instructions on deploying to a free Gradio app on Hugging Face Spaces.

Architecture

Topic Input → Framework Selection → Parallel Section Processing
                                          ↓
Complexity Assessment → Query Generation → Web Research → Analysis → Quality Check → Editor
                                          ↓
Self-Healing Loop (if needed) → Final Report Generation → Export (JSON/Markdown)

7 Research Sections (specific-idea framework):

• Problem/Pain Analysis
• Buyer/Budget Identification
• ROI Evidence
• Competitive Landscape
• Go-to-Market Channels
• Risk Assessment
• Defensibility Analysis

Tech Stack

• Agents: OpenAI Agents SDK
• Web Research: Serper + Playwright
• UI: Gradio with async streaming
• Architecture: Async/await Python with progress callbacks

---

Research depth meets execution speed.