DiliGenix Tableau

• 

  DiliGenix Dashboard

Inspiration

Market research analysts spend 3-4 weeks manually gathering competitive intelligence - crawling websites, reading reports, analyzing competitors, and synthesizing insights. We watched businesses pay $50,000+ for reports that were outdated by the time they were delivered.

What if AI could do this in 5 minutes?

We envisioned a world where any startup, investor, or business leader could type a company name and instantly receive comprehensive market intelligence with interactive Tableau dashboards - democratizing access to insights that were previously available only to enterprises with deep pockets.

What it does

DiliGenix Apex is an AI-powered competitive intelligence platform that automates the entire research workflow:

1. 🔍 Intelligent Web Research: Recursive AI agents search the web using DuckDuckGo, mine relevant URLs, and extract content from hundreds of sources
2. 🧠 Deep Analysis: AI synthesizes information into structured intelligence reports covering:
   • SWOT Analysis
   • PESTLE Analysis
     
   • Competitive Landscape
   • Market Positioning
   • Growth Opportunities
   • Risk Assessment
3. 📊 Tableau Cloud Publishing: Automatically generates .hyper files and publishes data directly to Tableau Cloud via REST API
4. 🎨 Interactive Dashboards: Creates ready-to-use visualizations showing research vectors, source attribution, analysis metrics, and timeline data
5. 📄 Publication-Ready Reports: Exports comprehensive Markdown reports with citations and structured sections

From "Tesla" to complete market intelligence report in under 5 minutes.

How we built it

Architecture:

• Frontend: PyQt5 desktop application with modern dark UI
• AI Engine: Ollama (local LLM) orchestrated through recursive sectional agents
• Data Pipeline:
  • Web scraping with trafilatura and BeautifulSoup4
  • Search integration via duckduckgo-search
  • Content extraction and synthesis
• Tableau Integration:
  • tableauhyperapi for native .hyper file generation
  • tableauserverclient for REST API publishing to Tableau Cloud
  • Unified Extract table with 4 data types (Research Vectors, URLs, Analysis, Metadata)
• Visualization: Plotly for standalone interactive HTML dashboards

Key Technical Innovations:

1. Recursive Sectional Agent Pattern: AI agents that spawn sub-agents for parallel research
2. Progressive Intelligence Gathering: Each query builds on previous findings
3. Real-time Progress Tracking: Live updates with query counts, URL mining, and processing status
4. Dual Export System: Both Tableau Cloud API publishing AND local CSV/HTML generation

Development Stack:

# Core Technologies
Python 3.14
PyQt5 (GUI Framework)
Ollama (AI/LLM Client)
Tableau APIs (Hyper + REST)

Built With

• duckduckgo-search
• hyper
• markdown
• ollama
• plotly
• pyqt
• python
• tableau
• tableau-hyper-api
• tableau-server-client