3. Data, Science & AI

Quick reference guide for common SQL patterns, including CTEs, window functions, indexing strategies, and pagination techniques.

Analyzes Claude Code session logs to extract insights on thinking blocks, tool usage, errors, and debug trajectories for improved development workflows.

Curates and restructures ML repositories for clean release, ensuring reproducibility while preserving evaluation data, paper assets, and minimal layout.

Provides vector-based memory storage and retrieval for LLMs using LanceDB and FastEmbed, enabling context-aware interactions and RAG capabilities.

High-performance web crawler utilizing Sidecar Execution Pattern for efficient and scalable data collection from web sources.

Handles safe data ingestion from S3 into bauplan using Write-Audit-Publish pattern, supporting parquet, CSV, and JSONL with quality checks.

Enables safe, read-only exploration of Bauplan lakehouse data via Python SDK, inspecting namespaces, tables, schemas, and exporting result sets.

Creates structured data pipeline projects with SQL and Python models, including DAG definitions and ETL workflow setup from scratch.

Enables local speech-to-text conversion using the Whisper CLI without API key requirements.

Batch-generates images via OpenAI API with random prompts and outputs an HTML gallery for display.

Transcribes audio files into text using OpenAI's Whisper API, enabling accurate speech-to-text conversion for various applications.

Provides current weather and forecast data without requiring an API key, enabling quick access to real-time weather information.

Enables offline text-to-speech synthesis using the sherpa-onnx library, eliminating cloud dependency for local speech generation.

Summarizes per-model usage and cost data for Codex and Claude from CodexBar CLI output, providing current model or full breakdown.

Command-line interface for instant AI-powered Q&A, text summarization, and content generation using Google's Gemini model.

Generates spectrograms and feature-panel visualizations from audio files using the songsee command-line interface for data analysis and audio processing.

Provides best practices for efficiently using the oracle CLI, including prompt management, file bundling, engine configuration, and session handling for AI workflows.

Prevents context window overflow in LLM workflows by processing large data via Map-Reduce, evidence contracts, and sub-agents.

Generates Entity-Relationship diagrams for ORE Studio database schema using PlantUML syntax.

Generates SQL table definitions, triggers, and data population scripts for ORE Studio database initialization.

Guides sociology researchers in computational text analysis using R/Python, covering topic modeling, sentiment analysis, and neural methods with validation.

Develops causal diagrams (DAGs) from social science research using R, Python, or Mermaid for publication-ready figures.

Guides sociological researchers through R-based statistical workflows for DiD, IV, and panel data analysis, producing publication-ready results.

Provides structured Stata-based statistical analysis guidance for sociology research, covering DiD, IV, matching, and panel methods for publication-ready results.