3. Data, Science & AI

Teaches foundational meta-analysis concepts including effect sizes, statistical models, and evidence synthesis for academic and research applications.

Optimizes token usage in Claude Code for cost efficiency, applying best practices in file handling, command execution, and model selection (Opus/Sonnet).

Provides coaching to clarify AI model training goals, diagnose issues, and recommend optimal approaches for fine-tuning and NLP tasks.

Creates, validates, and optimizes Galaxy workflows for scientific data analysis following IWC standards to ensure reproducibility and efficiency.

Provides state checkpointing, divergence detection, and confidence tracking for reliable multi-step reasoning and error recovery in AI systems.

Provides structured reference information on LLM fine-tuning techniques, model comparisons, architectures, and troubleshooting for AI practitioners.

Guide for building AI applications with Vercel AI SDK v6, covering LLM features like tool calling, agents, and structured output in TypeScript/JavaScript frameworks.

Provides a computational mathematics toolkit for algorithm analysis, discrete math, and probability based on MIT's curriculum, implemented in Python for Claude Code integration.

Provides foundational knowledge of bioinformatics data formats, sequencing technologies, and processing patterns for genomic data analysis and debugging.

Analyzes and optimizes Jupyter notebooks for scientific data analysis, emphasizing data filtering, visualization, and resource efficiency.

Guides users through LLM fine-tuning with coach-style workflow, supporting LoRA/QLoRA, SFT, and full version tracking for reproducible model training.

Provides Apple Foundation Models framework for on-device AI, enabling guided generation, tool calling, and streaming for local machine learning tasks.

Generates images from text prompts using Gemini 3 Pro AI model, supporting multiple aspect ratios and up to 4K resolution.

Pattern for building data collectors that fetch, transform, and deduplicate incident data from external APIs for efficient pipeline management.

Animates a single image into a video using fal.ai Veo 3.1, supporting up to 8 seconds with optional audio.

Tests and benchmarks DE algorithm variants (best, rand, pbest, current-to-best) for optimization performance and integration across store backends.

Provides instant access to AI model specifications, pricing, and capabilities from models.dev for comparison and SDK integration.

Edits existing images using text prompts via Gemini 3 Pro AI model, supporting multiple input images for context.

Provides search and query capabilities for a user's Goodreads library using a CSV export, answering questions about books, TBR, ratings, and reading status.

Automatically routes complex coding tasks to specialized AI agents via Roundtable MCP for multi-model verification and long-context analysis.

Delegates complex tasks to Codex AI via structured prompts, managing tmux sessions and logging for efficient AI-assisted development workflows.

Performs DEM lineament analysis to identify unmapped fault structures, aiding in earthquake cluster confirmation and fault mapping.

Calculates z-scores for speleothem geochemical data to identify anomalies and potential earthquake signals in cave isotope analysis.

Provides feedback and suggestions for improving AI context files (CLAUDE.md, context.md) to enhance AI coding assistant performance through context engineering.