Data Science & ML Claude 스킬

Jelly XLSX Specialist

Automates professional spreadsheet creation, data analysis, and financial modeling with industry-standard formatting and error-free formula calculation.

Marimo Reactive Notebooks

Develops and manages reactive Python notebooks that function as version-control-friendly scripts and deployable web apps.

HiPerGator Multi-Group Conda Manager

Configures and manages multiple Conda environment locations across different research group storage allocations on the UF HiPerGator supercomputer.

Scientific Project Data Separation

Organizes scientific research repositories by decoupling core code from experimental data and notebook outputs.

Windows CuPy NVRTC Fix

Resolves CuPy runtime compilation errors on Windows by correctly configuring CUDA NVRTC paths.

Veo 3.1 Image-to-Video Prompting

Generates high-fidelity video prompts for Google's Veo 3.1 model by analyzing images and applying professional cinematic formulas.

Channel Name Parsing Utility

Parses microscopy channel and marker names from KINTSUGI metadata files using automatic format detection.

LLM Evaluation Metrics

Standardizes LLM performance evaluation through automated datasets, A/B testing, and advanced LLM-as-judge patterns.

LLM Model Selection & Provider Management

Optimizes LLM performance and reliability through automated model selection, cost estimation, and intelligent fallback strategies.

Prompting Patterns

Implements standardized prompt engineering patterns to ensure reliable, secure, and high-performance LLM interactions.

LLM Application Architecture

Implements robust patterns for LLM-powered applications including async calls, streaming, token management, and resilient error handling.

Agent Orchestration Patterns

Implements robust multi-agent systems using Pydantic tool schemas, state management, and advanced orchestration patterns.

KINTSUGI GPU & Quality Priority

Enforces scientific-grade quality standards and mandatory GPU acceleration for KINTSUGI image processing pipelines.

Markov Trading Regime Features

Implements and debugs dynamic Hidden Markov Model (HMM) regime detection for Reinforcement Learning trading observations.

RAG Design Patterns

Implements high-performance Retrieval Augmented Generation (RAG) workflows including semantic chunking, vector database integration, and advanced reranking.

Vector Database Management

Facilitates seamless integration with vector databases for semantic search, retrieval-augmented generation (RAG), and high-dimensional embedding management.

Tile Snake Pattern Stitching

Reconstructs multiplex microscopy images by correctly ordering tiles acquired via snake or serpentine stage patterns.

BaSiC Caching Evaluation

Guides bioimaging optimization by advising against BaSiC caching and providing high-performance GPU-accelerated alternatives.

Mem0 Memory Optimization

Optimizes Mem0 performance through advanced query tuning, multi-layer caching strategies, and cost-reduction patterns.

ML Training Patterns

Streamlines machine learning model development with production-ready templates for classification, text generation, and parameter-efficient fine-tuning.

Python Performance Optimization Patterns

Optimizes Python codebases using verified, low-risk patterns for I/O, NumPy operations, and efficient directory scanning.

OpenRouter Model Routing Patterns

Implements intelligent model routing strategies to optimize for cost, performance, and reliability when building AI applications with OpenRouter.

ML Integration Helpers

Deploys machine learning models into full-stack applications using production-ready FastAPI endpoints, Next.js UI components, and Supabase schemas.

RAG Embedding Model Manager

Optimizes embedding model selection, configuration, and cost estimation for RAG pipelines.

Google File Search RAG

Implements managed Retrieval-Augmented Generation (RAG) using Google File Search and Gemini models for high-accuracy document retrieval and grounding.

Weaviate RAG Integration

Implements and manages Retrieval-Augmented Generation (RAG) systems using Weaviate vector databases for semantic search and document retrieval.

ML Training Validation Scripts

Ensures the integrity of machine learning training workflows by validating datasets, model checkpoints, and system dependencies.

ML Training Cost Calculator

Calculates and compares machine learning training and inference costs across major cloud GPU platforms like Modal, Lambda Labs, and RunPod.

RAG Chunking Strategies

Implements and benchmarks advanced document chunking strategies to optimize retrieval performance and context preservation in RAG pipelines.

Mem0 Memory Design Patterns

Architects scalable AI memory systems with optimized retention, storage backends, and multi-level context patterns.