3. Data, Science & AI

Automates real-time bidirectional synchronization between Aurora YAML schemas and Google Sheets for consistent data management.

Optimizes document handling and caching for Gemini AI workflows, enabling efficient token usage and multi-query strategies via File API integration.

Enables vector operations for particle physics data including deltaR, invariant mass, and vector boosting using scikit-hep vector and Awkward Array.

Provides prompt engineering guidance for GLM-4.7 model, optimizing user queries to enhance AI response quality and effectiveness.

Runs and interprets File API vs Inline benchmark for Gemini AI model performance, aiding optimization and document upload strategy decisions.

Enables querying ATLAS xAOD data via ServiceX and func_adl, with guidance for dataset selection, filtering, and delivery of physics data skims.

Removes backgrounds from images using FAL.ai's BiRefNet AI model, generating transparent PNGs for cutouts and subject extraction.

Executes mortgage underwriter CLI for loan document analysis, covering setup, execution, and result interpretation.

Provides advanced maximum likelihood phylogenetics with model selection, tree search algorithms, and GPU-accelerated performance optimization for evolutionary biology research.

Enables creation, filling, slicing, and plotting of 1D/2D/multi-axis histograms with scikit-hep hist library and mplhep conventions.

Validates LangGraph implementations for correctness, consistency, and integration readiness in AI agent workflows.

Manages foundational identity, response format, stack preferences, security protocols, and delegation patterns for the Personal AI Infrastructure (PAI).

Generates intelligent test cases using the Minion AI framework for automated software validation.

Optimizes Claude Code context window usage via JIT retrieval, progressive loading, and memory management for efficient large project workflows.

Proposes structured interpretations for vague user requests by restating, structuring, and committing to a next action in AI interactions.

Enables data exploration, transformation, and visualization to generate analytical reports and insights.

Configures Model Context Protocol (MCP) servers and plugins for Claude Code, enabling integration with external tools like GitHub, Brave Search, and SQLite.

Designs and refactors ML/RL pipelines for beam tracking, enforcing shape contracts and ensuring inference safety.

Extracts subtitles, frames, and metadata from YouTube videos for transcript, screenshot, and analysis when a URL is provided.

All-in-one Gemini 3 Pro client enabling text queries, file analysis (PDF, images, MP4), YouTube video processing, and image generation without API keys.

Extracts structured data from websites using Playwright, handling authentication, pagination, and dynamic SPAs for data collection.

Generates 2D chemical structure images and interactive 3D viewers from chemical names or SMILES, supporting PNG/SVG and 3Dmol.js formats.

Optimizes Alfred's learning system by analyzing session patterns and implementing adaptive learning to enhance performance and knowledge optimization.

Parses cloud charging pile communication protocol messages with CRC16 support, frame type handling, and error diagnosis, outputting standardized JSON data.