Data Science & ML Claudeスキル

Vector Database Master

Optimizes vector storage and retrieval strategies for high-performance AI applications and semantic search.

LLM Evaluation Metrics

Measures and optimizes Large Language Model performance through systematic quality frameworks, benchmarks, and hallucination detection.

Fine-Tuning AI Models

Adapts and optimizes Large Language Models using LoRA, QLoRA, and instruction tuning for domain-specific applications.

LLM Fundamentals & Architecture

Simplifies Large Language Model integration by providing expert guidance on transformer architecture, tokenization, and inference optimization.

Model Deployment

Deploys and optimizes large language models using production-grade frameworks like vLLM, TGI, and FastAPI.

Agent Frameworks & Orchestration

Develops autonomous AI agents and multi-agent systems using industry-standard frameworks like LangChain, CrewAI, and AutoGen.

RAG Implementation for LLMs

Builds robust Retrieval-Augmented Generation (RAG) systems using vector databases and semantic search to ground AI responses in proprietary data.

RAG Chunking Strategies

Implements optimized document splitting and processing workflows for Retrieval-Augmented Generation (RAG) systems.

Memvid Local RAG

Enables local semantic search and document indexing for PDF, DOCX, and XLSX files directly within Claude Code.

OpenAI Integration Specialist

Integrates sophisticated OpenAI API features including streaming chatbots, content generation, and rate-limited API routes for production-ready AI applications.

Subgroup Analysis Pro

Conducts rigorous statistical subgroup analyses to identify effect moderation and data heterogeneity across diverse research populations.

Incentive Prompting Optimizer

Optimizes AI outputs using research-backed prompting techniques to increase response quality and accuracy by up to 115%.

Data Visualization

Generates publication-quality scientific figures and data visualizations following academic design standards.

Research Pre-Registration

Automates the creation of comprehensive pre-registration documents to ensure research transparency and prevent questionable scientific practices.

PRISMA Flow Diagram Generator

Generates standardized PRISMA 2020 flow diagrams to document the study selection process in systematic literature reviews.

Literature Gap Identifier

Identifies and prioritizes research gaps from systematic literature reviews to justify new studies and grant proposals.

Blinding Procedures for Research

Implements rigorous blinding protocols to minimize bias and ensure objectivity in experimental studies and clinical trials.

PolicyEngine Aggregation Patterns

Implements standardized patterns for summing tax and benefit variables across entities using the adds attribute and add() function.

Incentive Prompting & AI Optimization

Enhances AI response quality by 45-115% using research-backed techniques like monetary framing, expert personas, and step-by-step reasoning.

MLOps & Production Machine Learning

Automates machine learning lifecycles through experiment tracking, model versioning, and production-grade deployment pipelines.

Effect Size Calculator

Calculates and interprets standardized effect sizes to quantify the magnitude of research findings beyond simple p-values.

Synthesis Matrix

Generates structured evidence synthesis matrices to organize and compare research data for systematic reviews.

SAP RPT-1 Predictor

Leverages SAP's open-source tabular foundation model to perform predictive analytics on structured business data without model training.

Data Science & Sparsification

Implements L0 regularization and intelligent sampling techniques to optimize neural network sparsification and survey data calibration.

MQL5 Python Workspace

Automates the translation of MetaTrader 5 (MQL5) indicators into validated Python implementations for algorithmic trading.

LLM Evaluation & Benchmarking

Implements comprehensive evaluation frameworks for LLM applications using automated metrics, LLM-as-Judge patterns, and human feedback loops.

ReasoningBank Intelligence

Implements adaptive learning and meta-cognitive capabilities for AI agents to optimize strategies based on historical experience.

ReasoningBank with AgentDB

Implements high-performance adaptive learning and memory distillation for AI agents using the AgentDB vector backend.

HoloViz Plotting Fundamentals

Simplifies the creation of interactive, publication-quality visualizations using hvPlot and HoloViews within the HoloViz ecosystem.

HoloViz Colormaps & Styling

Applies perceptually uniform colormaps and accessible visual styling to data visualizations using Colorcet and the HoloViz ecosystem.