voltagent-data-ai — плагин для Claude Code: субагенты для данных и ML

# Data & AI Subagents

Data & AI subagents are your specialists in the world of data engineering, machine learning, and artificial intelligence. These experts handle everything from building robust data pipelines to training sophisticated ML models, from optimizing databases to deploying AI systems at scale. They bridge the gap between raw data and intelligent applications, ensuring your data-driven solutions are efficient, scalable, and impactful.

## When to Use Data & AI Subagents

Use these subagents when you need to:
- **Build data pipelines** for ETL/ELT workflows
- **Train machine learning models** for predictions and insights
- **Design AI systems** for production deployment
- **Optimize database performance** at scale
- **Implement NLP solutions** for text processing
- **Create computer vision** applications
- **Deploy ML models** with MLOps best practices
- **Analyze data** for business insights

## Available Subagents

### [**ai-engineer**](ai-engineer.md) - AI system design and deployment expert
AI systems specialist building production-ready artificial intelligence solutions. Masters model deployment, scaling, and integration. Bridges the gap between AI research and real-world applications.

**Use when:** Deploying AI models to production, designing AI system architectures, integrating AI into applications, scaling AI services, or implementing AI pipelines.

### [**data-analyst**](data-analyst.md) - Data insights and visualization specialist
Analytics expert transforming data into actionable insights. Masters statistical analysis, data visualization, and business intelligence tools. Tells compelling stories with data.

**Use when:** Analyzing business data, creating dashboards, performing statistical analysis, building reports, or discovering data insights.

### [**data-engineer**](data-engineer.md) - Data pipeline architect
Data infrastructure specialist building scalable data pipelines. Expert in ETL/ELT processes, data warehousing, and streaming architectures. Ensures data flows reliably from source to insight.

**Use when:** Building data pipelines, designing data architectures, implementing ETL processes, setting up data warehouses, or handling big data processing.

### [**data-scientist**](data-scientist.md) - Analytics and insights expert
Data science practitioner combining statistics, machine learning, and domain expertise. Masters predictive modeling, experimentation, and advanced analytics. Extracts value from complex datasets.

**Use when:** Building predictive models, conducting experiments, performing advanced analytics, developing ML algorithms, or solving complex data problems.

### [**database-optimizer**](database-optimizer.md) - Database performance specialist
Database performance expert ensuring queries run at lightning speed. Masters indexing strategies, query optimization, and database tuning. Makes databases perform at their peak.

**Use when:** Optimizing slow queries, designing efficient schemas, implementing indexing strategies, tuning database performance, or scaling databases.

### [**llm-architect**](llm-architect.md) - Large language model architect
LLM specialist designing and deploying large language model solutions. Expert in prompt engineering, fine-tuning, and LLM applications. Harnesses the power of modern language models.

**Use when:** Implementing LLM solutions, designing prompt strategies, fine-tuning models, building chatbots, or creating AI-powered applications.

### [**machine-learning-engineer**](machine-learning-engineer.md) - Machine learning systems expert
ML engineering specialist building end-to-end machine learning systems. Masters the entire ML lifecycle from data to deployment. Ensures models work reliably in production.

**Use when:** Building ML pipelines, implementing ML systems, deploying models, creating ML infrastructure, or productionizing ML solutions.

### [**ml-engineer**](ml-engineer.md) - Machine learning specialist
Machine learning expert developing and optimizing ML models. Proficient in various algorithms, frameworks, and techniques. Solves complex problems with machine learning.

**Use when:** Training ML models, selecting algorithms, optimizing model performance, implementing ML solutions, or experimenting with new techniques.

### [**mlops-engineer**](mlops-engineer.md) - MLOps and model deployment expert
MLOps specialist ensuring smooth ML model deployment and operations. Masters CI/CD for ML, model monitoring, and versioning. Brings DevOps practices to machine learning.

**Use when:** Setting up ML pipelines, implementing model monitoring, automating ML workflows, managing model versions, or establishing MLOps practices.

### [**nlp-engineer**](nlp-engineer.md) - Natural language processing expert
NLP specialist building systems that understand and generate human language. Expert in text processing, language models, and linguistic analysis. Makes machines understand text.

**Use when:** Building text processing systems, implementing chatbots, analyzing sentiment, extracting information from text, or developing language understanding features.

### [**postgres-pro**](postgres-pro.md) - PostgreSQL database expert
PostgreSQL specialist mastering advanced features and optimizations. Expert in complex queries, performance tuning, and PostgreSQL-specific capabilities. Unlocks PostgreSQL's full potential.

**Use when:** Working with PostgreSQL, optimizing Postgres queries, implementing advanced features, designing PostgreSQL schemas, or troubleshooting Postgres issues.

### [**prompt-engineer**](prompt-engineer.md) - Prompt optimization specialist
Prompt engineering expert crafting effective prompts for AI models. Masters prompt design, testing, and optimization. Maximizes AI model performance through strategic prompting.

**Use when:** Designing prompts for LLMs, optimizing AI responses, implementing prompt strategies, testing prompt effectiveness, or building prompt-based applications.

### [**reinforcement-learning-engineer**](reinforcement-learning-engineer.md) - Reinforcement learning and agent training expert
RL specialist designing environments, shaping rewards, and training agents with modern algorithms like PPO, SAC, and DQN. Masters sim-to-real transfer, multi-agent systems, and safe exploration strategies for robotics, gaming, and autonomous decision-making.

**Use when:** Designing RL environments, training game AI or robotics agents, implementing policy gradient methods, optimizing reward functions, or deploying autonomous decision-making systems.

## Quick Selection Guide

| If you need to... | Use this subagent |
|-------------------|-------------------|
| Deploy AI systems | **ai-engineer** |
| Analyze business data | **data-analyst** |
| Build data pipelines | **data-engineer** |
| Create ML models | **data-scientist** |
| Optimize databases | **database-optimizer** |
| Work with LLMs | **llm-architect** |
| Build ML systems | **machine-learning-engineer** |
| Train ML models | **ml-engineer** |
| Deploy ML models | **mlops-engineer** |
| Process text data | **nlp-engineer** |
| Optimize PostgreSQL | **postgres-pro** |
| Design AI prompts | **prompt-engineer** |
| Train RL agents | **reinforcement-learning-engineer** |

## Common Data & AI Patterns

**End-to-End ML System:**
- **data-engineer** for data pipeline
- **data-scientist** for model development
- **ml-engineer** for model optimization
- **mlops-engineer** for deployment

**AI Application:**
- **llm-architect** for LLM integration
- **prompt-engineer** for prompt optimization
- **ai-engineer** for system design
- **nlp-engineer** for text processing

**Data Platform:**
- **data-engineer** for infrastructure
- **database-optimizer** for performance
- **postgres-pro** for PostgreSQL
- **data-analyst** for insights

**Production ML:**
- **machine-learning-engineer** for ML systems
- **mlops-engineer** for operations
- **ai-engineer** for deployment
- **data-engineer** for data flow

**RL Systems:**
- **reinforcement-learning-engineer** for agent training
- **ml-engineer** for training infrastructure
- **mlops-engineer** for model deployment
- **ai-engineer** for system integration

## Getting Started

1. **Define your data/AI objectives** clearly
2. **Assess your data landscape** and requirements
3. **Choose appropriate specialists** for your needs
4. **Provide data context** and constraints
5. **Follow best practices** for implementation

## Best Practices

- **Start with data quality:** Good models need good data
- **Iterate quickly:** ML is experimental by nature
- **Monitor everything:** Models drift, data changes
- **Version control:** Track data, code, and models
- **Document thoroughly:** ML systems are complex
- **Test rigorously:** Validate models before production
- **Scale gradually:** Start small, prove value
- **Stay ethical:** Consider AI's impact

Choose your data & AI specialist and unlock the power of your data today!