3. Data, Science & AI

Specializes in database management for PostgreSQL, MongoDB, Redis, and Oracle, implementing advanced data patterns for modern application development.

Continuous learning system that extracts, stores, and reuses code patterns from Claude Code sessions to enhance AI-driven development workflows.

Provides foundational principles for AI-powered development workflows, including agent catalog, token optimization, and modular architecture.

Guides implementation of Vercel AI SDK v5/v6 features including useChat, tool(), streamText, and agent patterns for Next.js AI applications.

Canonical kit for building Claude Code extensions with agent patterns, plugins, and configuration settings.

Enables parallel subagent execution for multi-perspective analysis, adversarial review, and coordinated task delegation in AI agent systems.

An EDD framework that treats evaluation metrics as unit tests in AI development, ensuring model reliability and performance through systematic testing.

Orchestrates multi-stage command triage and decision routing through three AI strategists using LangGraph for efficient LLM-based agent workflows.

Utilizes Claude Opus 4.5's extended thinking for deep reasoning, complex analysis, and multi-step synthesis to deliver thorough, well-considered responses.

Provides patterns and best practices for coordinating multiple AI agents, task decomposition, and workflow orchestration in agent-based systems.

Manages enterprise AI session context with token budget optimization and state persistence for cost-effective LLM interactions.

Provides geocoding and location resolution for OpenWeatherMap API to fetch weather data, forecasts, and air quality.

Optimizes Suno AI prompts to produce commercial-grade music with precise creative direction and technical quality.

Creates a new Dagster workspace to manage multiple data projects and orchestrate data pipelines in a unified environment.

Expert guidance for building and extending Dagster data orchestration projects, covering assets, schedules, sensors, and ETL patterns.

Creates production-ready data pipelines and workflows using Dagster with best practices, testing, and validation.

Comprehensive index of Dagster integrations organized by data engineering categories for efficient tool discovery and selection.

Real-time tracking of Pittsburgh snow plow locations and snow response status using the City of Pittsburgh's official dashboard data.

Guides setup and management of research-judge agents for batch evaluation, covering directory structure, parallel execution, and result aggregation.

Standardizes folder structure and templates for organizing data science experiments, ensuring consistency and reproducibility.

Validates AI prompt files against reference guidelines to ensure compliance and quality for LLM applications.

Enables distributed data processing by splitting tasks (map) and aggregating results (reduce) for efficient large-scale analysis and transformations.

Builds statistical arbitrage trading systems using rigorous data science methods for alpha research and signal extraction in quantitative finance.

Assists in designing reliable data infrastructure for databases, streaming systems, and data pipelines using Kleppmann's data-intensive application principles.