Data Science & ML Claude 技能

Advanced LLM Evaluation

Implements production-grade LLM-as-a-Judge techniques for evaluating AI outputs through rigorous scoring, pairwise comparisons, and bias mitigation.

Phoenix Advanced Financial Forecasting

Powers probabilistic fund performance modeling through graduation rates, multi-MOIC analysis, and Monte Carlo simulations.

Phoenix Reserves Optimizer

Optimizes venture capital reserve allocations and follow-on investment strategies within the Phoenix fund model.

Alpha Vantage Financial Data

Integrates Alpha Vantage APIs to fetch real-time and historical market data, technical indicators, and financial fundamentals.

Awkward Array 2.0 Development

Streamlines the development of Python workflows using Awkward Array for jagged, nested, and record-based data structures.

Similarity Search Patterns

Implement efficient vector-based similarity search, semantic retrieval, and RAG patterns across major vector databases.

Experiment Advisor

Proposes structured, data-driven experiment plans by analyzing historical training reports, logs, and research goals.

Context Compression Strategies

Optimizes long-running AI agent sessions by implementing structured context compression to maintain technical accuracy and memory efficiency.

FinnHub Financial Data

Integrates Claude with the FinnHub API to retrieve real-time stock quotes, fundamental data, crypto prices, and market news.

Senior Prompt Engineer & AI Architect

Implements production-grade prompt engineering patterns, RAG optimization, and agentic system architectures for advanced AI products.

xAI Agent Tools for Grok

Empowers autonomous AI agents with real-time X (Twitter) search, web search, and sandboxed Python code execution capabilities.

LLM Evaluation & Benchmarking

Implements comprehensive evaluation frameworks for LLM applications using automated metrics, human-in-the-loop feedback, and A/B testing.

Vaex Big Data Science

Processes and analyzes billion-row tabular datasets using lazy, out-of-core DataFrame operations without exceeding available RAM.

Data Storytelling & Analytics Narratives

Transforms raw data into persuasive narratives and visualizations for business stakeholders and executive presentations.

Spark Optimization

Optimizes Apache Spark performance through advanced partitioning, memory tuning, and shuffle management strategies.

Embedding Strategies & RAG Optimization

Implements and optimizes text embedding models and chunking strategies for high-performance semantic search and RAG applications.

RAG Implementation Expert

Build robust Retrieval-Augmented Generation systems for LLM applications using vector databases and semantic search.

LLM Gateway for Rails

Builds a unified abstraction layer for integrating and switching between multiple LLM providers in Rails applications.

Vector Index Tuning

Optimizes vector database performance by tuning HNSW parameters, quantization strategies, and memory usage for high-scale search applications.

Derivation Checker

Verifies mathematical derivations step-by-step by checking algebra, index consistency, and dimensional accuracy.

ML Experiment Logger

Logs machine learning experiments with hyperparameters, metrics, and visualizations for systematic research tracking.

LangChain Architecture

Architects sophisticated LLM applications using agents, memory, and tool integration within the LangChain framework.

ML Pipeline Workflow

Orchestrates end-to-end MLOps pipelines from data preparation and model training to production deployment and monitoring.

ServiceX for ATLAS xAOD

Generates and optimizes ServiceX queries in func_adl for analyzing ATLAS xAOD high-energy physics data.

Python Experiment Conventions

Standardizes Python experiment layouts, stage entrypoints, and asset handling for consistent data science workflows.

Quantitative Backtesting Frameworks

Builds robust, production-grade backtesting systems for trading strategies while eliminating common statistical biases.

Reactome Biological Pathway Analysis

Queries the Reactome REST API to perform pathway enrichment analysis, gene mapping, and systems biology research directly within Claude.

Quantitative Risk Metrics Calculation

Calculates comprehensive portfolio risk metrics like VaR, CVaR, and Sharpe ratios to monitor and manage financial exposure.

AI Multimodal Processing

Integrates Google Gemini's advanced multimodal capabilities to process, analyze, and generate audio, video, images, and documents directly within your development workflow.

Prompt Engineering Patterns

Master advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability in production.