RivalForge AI
RivalForge automates competitive intelligence in four stages: 1. **Discover** — finds competitor URLs via Perplexity Sonar Pro or Exa AI 2. **Extract** — crawls competitor sites with Firecrawl for pricing, features, tech stack, and feedback 3. **Compare** — builds a side-by-side comparison table across all rivals 4. **Analyze** — generates a strategic report with market gaps, weaknesses, and differentiation opportunities
RivalForge AI
Competitor intelligence agent team by Sharad Khare.
RivalForge AI is a multi-agent competitor analysis platform that discovers rival companies, extracts structured website data with Firecrawl, and generates comparison tables plus strategic opportunity reports.
Built and maintained by Sharad Khare — AI strategist, full-stack developer, and creator of practical agent workflows.
What this project does
RivalForge automates competitive intelligence in four stages:
1. Discover — finds competitor URLs via Perplexity Sonar Pro or Exa AI
2. Extract — crawls competitor sites with Firecrawl for pricing, features, tech stack, and feedback
3. Compare — builds a side-by-side comparison table across all rivals
4. Analyze — generates a strategic report with market gaps, weaknesses, and differentiation opportunities
Why Sharad Khare built RivalForge AI
Competitive research is slow and fragmented — finding rivals, scraping sites, and synthesizing insights takes hours. RivalForge demonstrates how a coordinated agent team can compress that workflow into a single Streamlit experience.
How it works
Company URL or Description
│
▼
Competitor Discovery (Perplexity or Exa)
└── 3 competitor URLs
│
▼
Firecrawl Extraction (per competitor)
├── Pricing & plans
├── Key features
├── Tech stack
├── Marketing focus
└── Customer feedback
│
▼
Comparison Table (pandas)
│
▼
Strategic Analysis Agent (OpenAI)
└── Market gaps, positioning, and growth opportunities
Agent roles
| Stage | Tool/Model | Output |
|-------|-----------|--------|
| Discovery | Perplexity Sonar Pro or Exa AI | Competitor URLs |
| Extraction | Firecrawl + structured schema | JSON per competitor |
| Comparison | pandas DataFrame | Side-by-side table |
| Analysis | OpenAI GPT-4o | Strategic opportunity report |
Features
- Dual search endpoints: Perplexity Sonar Pro or Exa AI
- Firecrawl-powered website extraction with Pydantic schemas
- Structured competitor data: pricing, features, tech stack, marketing, feedback
- Interactive comparison table in Streamlit
- AI-generated strategic analysis with actionable recommendations
- Works with company URL, description, or both
Quick start
cd rivalforge-ai
pip install -r requirements.txt
streamlit run app.py
Required API keys
- OpenAI API key — analysis and Exa-based discovery
- Firecrawl API key — competitor website extraction
- Perplexity API key — if using Perplexity Sonar Pro search
- Exa API key — if using Exa AI search
Example inputs
- URL:
https://yourcompany.com+ Description:B2B project management SaaS - Description only:
AI-powered customer support chatbot for e-commerce - URL only:
https://notion.so
Project structure
rivalforge-ai/
├── app.py
├── rivalforge/
│ ├── config.py
│ ├── schemas.py
│ ├── agents.py
│ ├── discovery.py
│ ├── extraction.py
│ └── reports.py
├── requirements.txt
├── pyproject.toml
└── README.md
Who this is for
- Startup founders doing competitive landscape research
- Product managers benchmarking rival features and pricing
- Marketing teams analyzing competitor positioning
- Developers learning multi-agent + web extraction patterns
Use cases
- Discover top 3 competitors from a company URL or description
- Extract structured pricing and feature data from rival websites
- Build comparison tables for stakeholder presentations
- Generate strategic reports highlighting differentiation opportunities
Author
- Website: sharadkhare.in
License
MIT © Sharad Khare
Loading file tree…
Select a file to preview its source.
cd rivalforge-ai
pip install -r requirements.txt
streamlit run app.py