OpenProject Haystack

AI-Powered Project Management

---

Table of Contents

• 🎯 Executive Summary
• 🔒 Responsible AI Principles
• 🏗️ System Architecture
• ⚡ Quick Start
• 🔧 Technical Documentation
• 📚 Detailed Documentation

---

🎯 Executive Summary

OpenProject Haystack is an AI-powered project management enhancement system that delivers automated compliance, proactive risk management, and data-driven decision support. The system transforms traditional project management workflows into intelligent, automated processes that ensure PMFlex compliance while significantly reducing manual overhead.

Key Business Benefits

Capability	Traditional Approach	Our AI Solution	Business Impact
Status Reports	4-8 hours manual work	2 minutes automated	80% time savings
Issue Detection	Reactive (after problems occur)	Predictive (2-4 weeks early)	40% reduction in overruns
Compliance	Manual review and correction	Automatic PMFlex adherence	95% reduction in audit issues
Quality	Inconsistent, human error prone	AI-powered, standardized	90% reduction in review time

Strategic Value Propositions

• 🎯 100% PMFlex Compliance: Automatic adherence to German federal project management standards
• 📈 Proactive Risk Management: Early detection of project health issues before they become critical
• 💼 Data-Driven Decision Support: Real-time project health analytics and predictive insights
• ⚡ Operational Efficiency: 80% time savings through automated report generation and workflows

📈 View Complete Executive Summary - Full business case, ROI analysis, and strategic benefits

---

🔒 Responsible AI Principles

OpenProject Haystack follows strict ethical AI guidelines to ensure responsible deployment in government environments:

Core Principles

🏠 Self-Hosting & Data Sovereignty

• All AI models run locally on your infrastructure
• Complete control over data processing and storage
• No external data transmission or cloud dependencies
• Minimizes security risks and ensures compliance

👤 Human Decision Authority

• AI serves as decision support, never autonomous decision-making
• Transparent recommendations with clear explanations
• Human oversight required for all critical decisions
• Clear accountability chains maintained

🔐 Privacy & GDPR Compliance

• Strict data isolation between projects and tenants
• Minimal data collection principles
• Automated data deletion routines
• Complete audit trails and documentation

🌱 Sustainability & Ethics

• Energy-efficient model selection
• Ethical training data sourcing
• Regular bias testing and mitigation
• Cultural fairness and inclusivity

🔒 Read Full AI Governance Guidelines - Comprehensive AI governance and compliance framework

---

🏗️ System Architecture

Complete System Overview

graph TB
    %% External Systems
    OP[🏢 OpenProject Instance<br/>Project Data Source]
    USER[👤 Project Manager<br/>End User]
    
    %% PMFlex Knowledge Base
    PM[📚 PMFlex.one<br/>German Federal Standards<br/>• Handbooks<br/>• Templates<br/>• Guidelines]
    
    %% Haystack System Core
    subgraph "🤖 OpenProject Haystack System"
        API[🔌 API Layer<br/>FastAPI Application<br/>• OpenAI Compatible<br/>• Project Endpoints]
        
        subgraph "🧠 AI Intelligence Core"
            RAG[🔍 RAG Pipeline<br/>Document Retrieval<br/>• Vector Store<br/>• Context Enhancement]
            LLM[⚡ Ollama LLM<br/>German Language<br/>• Report Generation<br/>• Hint Creation]
        end
        
        subgraph "📊 Analysis Engine"
            ANALYZER[🔬 Project Analyzer<br/>• 10 Health Checks<br/>• Risk Assessment<br/>• Performance Metrics]
        end
    end
    
    %% Data Flow
    USER --> API
    API --> OP
    OP --> ANALYZER
    ANALYZER --> RAG
    PM --> RAG
    RAG --> LLM
    LLM --> API
    API --> USER
    
    %% Styling
    classDef external fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
    classDef knowledge fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
    classDef core fill:#e8f5e8,stroke:#388e3c,stroke-width:2px
    classDef ai fill:#fff3e0,stroke:#f57c00,stroke-width:2px
    classDef analysis fill:#fce4ec,stroke:#c2185b,stroke-width:2px
    
    class OP,USER external
    class PM knowledge
    class API core
    class RAG,LLM ai
    class ANALYZER analysis

Loading

Key Components

Component	Primary Function	Business Value
PMFlex.one	German federal compliance standards	Ensures 100% regulatory adherence
API Layer	Interface for all system interactions	Seamless integration with existing tools
RAG System	Context-aware document retrieval	Intelligent, compliance-driven responses
Ollama LLM	German language AI generation	Professional, accurate content creation
Analysis Engine	Real-time project health monitoring	Proactive risk identification and management

Smart Report Generation Workflow

flowchart LR
    A[Project Manager Request] --> B[🔄 Automated Data Collection]
    B --> C[📊 AI Analysis & Insights]
    C --> D[📋 PMFlex Standards Integration]
    D --> E[📄 Professional German Report]
    
    B1[⏱️ 15 minutes] --> B
    C1[🎯 100% Accuracy] --> C
    D1[✅ Full Compliance] --> D
    E1[📈 Executive Ready] --> E
    
    style A fill:#e3f2fd
    style E fill:#c8e6c9
    style B fill:#fff3e0
    style C fill:#f3e5f5
    style D fill:#e8f5e8

Loading

🔄 See Detailed Workflow Documentation - Complete technical workflows and system diagrams

---

⚡ Quick Start

Development with Docker Compose

# Clone the repository
git clone https://github.com/opf/openproject-haystack.git
cd openproject-haystack

# Start the services (models will be automatically installed)
docker compose up --build

# The API will be available at:
# - Health check: http://localhost:8000/health
# - API docs: http://localhost:8000/docs
# - OpenAI compatible: http://localhost:8000/v1/chat/completions

Core Capabilities

📊 Project Status Reports

curl -X POST "http://localhost:8000/generate-project-status-report" \
  -H "Content-Type: application/json" \
  -d '{
    "project": {"id": 1, "type": "agile"},
    "openproject": {
      "base_url": "https://your-openproject.com",
      "user_token": "your-api-key"
    }
  }'

💡 Project Management Hints

curl -X POST "http://localhost:8000/project-management-hints" \
  -H "Content-Type: application/json" \
  -d '{
    "project": {"id": 1, "type": "agile"},
    "openproject": {
      "base_url": "https://your-openproject.com",
      "user_token": "your-api-key"
    }
  }'

🤖 OpenAI-Compatible Chat

curl -X POST "http://localhost:8000/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistral:latest",
    "messages": [
      {"role": "user", "content": "Help me with project management"}
    ]
  }'

---

🔧 Technical Documentation

Stack & Architecture

• Python 3.11 - Modern Python with type hints
• FastAPI - High-performance API framework
• Haystack AI - Advanced NLP pipeline management
• Ollama - Local LLM inference with German language support
• Docker & Docker Compose - Reproducible deployment environments
• RAG System - Context-aware document retrieval and enhancement

Project Structure

openproject-haystack/
├── README.md                    # This comprehensive documentation
├── EXECUTIVE_SUMMARY.md         # Business case and strategic benefits
├── LEITFADEN_VERANTWORTUNGSVOLLE_KI_OPENPROJECT.md  # AI governance
├── WORKFLOW_DIAGRAMS_AND_SUMMARIES.md  # Technical workflows
├── requirements.txt             # Python dependencies
├── Dockerfile                   # Container configuration
├── docker-compose.yml          # Multi-service orchestration
├── config/
│   └── settings.py             # Environment configuration
├── src/
│   ├── main.py                 # FastAPI application entry point
│   ├── api/
│   │   └── routes.py           # API endpoints (OpenAI + custom)
│   ├── pipelines/
│   │   ├── generation.py       # LLM generation pipelines
│   │   └── rag_pipeline.py     # RAG system implementation
│   ├── services/
│   │   ├── openproject_client.py  # OpenProject API integration
│   │   ├── document_processor.py  # Document processing
│   │   └── vector_store.py     # Vector database management
│   ├── models/
│   │   └── schemas.py          # Pydantic models and schemas
│   └── templates/
│       └── report_templates.py # Report generation templates
├── documents/
│   └── pmflex/                 # PMFlex compliance documents
└── tests/                      # Test suite

API Endpoints

Project Management Endpoints

• POST /generate-project-status-report - Generate comprehensive project status reports
• POST /project-management-hints - Get AI-powered project management recommendations
• GET /health - System health check

OpenAI-Compatible Endpoints

• POST /v1/chat/completions - Chat completion (OpenAI-compatible)
• GET /v1/models - List available models
• GET /v1/models/{model_id} - Get specific model information

RAG System Management

• POST /rag/initialize - Initialize RAG system with PMFlex documents
• GET /rag/status - Get RAG system status and statistics
• POST /rag/refresh - Refresh document index
• POST /rag/search - Search PMFlex documents

Configuration

Environment variables for customization:

# Core Configuration
OLLAMA_URL=http://ollama:11434
OLLAMA_MODEL=mistral:latest
API_HOST=0.0.0.0
API_PORT=8000

# Model Management
MODELS_TO_PULL=mistral:latest,llama2:7b
REQUIRED_MODELS=mistral:latest

# Generation Parameters
GENERATION_NUM_PREDICT=1000
GENERATION_TEMPERATURE=0.7

# Logging
LOG_LEVEL=INFO
LOG_FORMAT=%(asctime)s - %(name)s - %(levelname)s - %(message)s

# RAG System
RAG_CHUNK_SIZE=500
RAG_CHUNK_OVERLAP=50
RAG_MAX_CHUNKS=5

Testing

# Test OpenAI API compatibility
python test_openai_api.py

# Test project status report generation
python test_project_status_report.py

# Test project management hints
python test_project_hints.py

# Test RAG system
python test_haystack_api.py

Integration Examples

With OpenAI Python Client

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/v1",
    api_key="dummy-key"
)

response = client.chat.completions.create(
    model="mistral:latest",
    messages=[
        {"role": "system", "content": "You are a project management assistant."},
        {"role": "user", "content": "How can I improve my project timeline?"}
    ]
)

With LangChain

from langchain.chat_models import ChatOpenAI

llm = ChatOpenAI(
    openai_api_base="http://localhost:8000/v1",
    openai_api_key="dummy-key",
    model_name="mistral:latest"
)

Features

• ✅ AI-Powered Project Analysis: 10 automated health checks and risk assessments
• ✅ PMFlex Compliance: Automatic adherence to German federal standards
• ✅ German Language Support: Native German report generation and hints
• ✅ OpenAI API Compatibility: Drop-in replacement for OpenAI services
• ✅ Self-Hosted Security: Complete data sovereignty and privacy
• ✅ RAG-Enhanced Context: Intelligent document retrieval and context enhancement
• ✅ Real-Time Integration: Live OpenProject data analysis
• ✅ Professional Reporting: Executive-ready status reports and insights
• ✅ Predictive Analytics: Early warning system for project issues
• ✅ Automated Workflows: Streamlined project management processes

---

📚 Detailed Documentation

📋 Core Documentation

• Executive Summary - Complete business case, ROI analysis, strategic benefits, and implementation roadmap
• Responsible AI Guidelines - Comprehensive AI governance, ethical principles, and compliance framework
• Workflow Diagrams & Summaries - Technical workflows, system architecture, and detailed process documentation

🔧 Technical Documentation

• RAG System Documentation - RAG implementation details, document processing, and vector store management
• Deployment Guide - Production deployment instructions, scaling, and infrastructure requirements
• Project Hints API - API documentation for project management hints and recommendations

📊 Feature Documentation

• BlockNote AI Integration - BlockNote editor integration and streaming capabilities
• Hint Generation Enhancements - Advanced hint generation and optimization
• Memory Requirements - System requirements and performance optimization

🧪 Testing & Development

• Local Ollama Testing - Local development and testing procedures
• Mixtral Migration - Model migration and optimization guide
• Embedding Persistence - Document embedding and persistence strategies

---

🚀 Getting Started by Role

👔 For Executives

1. Read the Executive Summary for business benefits and ROI
2. Review the Responsible AI Guidelines for governance
3. Understand the strategic value proposition and competitive advantages

🔒 For Compliance Teams

1. Study the Responsible AI Guidelines for GDPR compliance
2. Review data sovereignty and self-hosting principles
3. Understand audit trails and documentation requirements

👨‍💻 For Developers

1. Follow the Quick Start guide for immediate setup
2. Explore the Technical Documentation for implementation details
3. Review the API Endpoints for integration options

📊 For Project Managers

1. Understand the core capabilities and business benefits
2. Test the project status report and hints generation features
3. Review the Project Hints API for practical usage

---

📞 Support & Community

• GitHub Issues: Report bugs and request features
• Documentation: Comprehensive guides and API documentation
• Community: Join discussions and share experiences
• Enterprise Support: Contact for enterprise deployment assistance

---

📜 License

This project is licensed under the GNU General Public License v3.0. See the LICENSE file for details.

---

OpenProject Haystack - Transforming project management through responsible AI innovation