3. Data, Science & AI

Provides exact symbolic mathematical computations for algebra, calculus, linear algebra, and more using SymPy.

Creates and manages scenario-based seed data bundles for PropertyWebBuilder, enabling pre-configured tenant website setups with property content.

Enables building machine learning and AI applications in Rust with support for model training, inference, and neural networks using Rust-specific frameworks.

Builds AI applications using LLMs, including prompt engineering, RAG, and integration for chatbots and automation features.

Enables creation, editing, and analysis of spreadsheets including data processing, formula recalculations, and visualization for data-driven tasks.

Enables Gemini model integration through OpenAI-compatible API with configuration for summary, vision, and response formatting.

Retrieves historical context from Slack memory storage to inform AI responses with project status, user preferences, and meeting decisions.

Automates AI memory cleanup by detecting duplicates, correcting misclassifications, and removing stale data to maintain efficient data organization.

Specializes in optimizing data formats (JSON, YAML, TOON) for LLM transmission, high-performance serialization, and schema validation in modern applications.

Specializes in R 4.4+ development for data analysis pipelines, visualizations using tidyverse and ggplot2, and Shiny application development.

Manages enterprise user sessions and context with token budget optimization for AI applications, ensuring state persistence across interactions.

Database specialist for PostgreSQL, MongoDB, Redis, and Oracle with advanced data patterns for modern applications.

Enhances output accuracy for non-reasoning LLMs using prompt repetition methodology.

Provides CLI-based knowledge storage and retrieval using AI DevKit's memory service, replacing MCP for AI workflow integration.

Optimizes user-provided prompts for AI models to improve response quality and relevance. Invoked immediately upon request without prior analysis.

Converts arXiv papers to LLM-optimized markdown from IDs, URLs, or PDF folders for AI processing.

Generates diverse solution options from multiple reasoning perspectives for a defined problem using an AI-driven workflow without preliminary exploration.

Provides immediate structured reasoning for open-ended analytical questions via Python script orchestration without preliminary exploration.

Validates data consistency in MikoPBX SQLite databases after REST API operations, checking foreign keys and CDR records for integrity.

Guides users in researching SQL syntax and behavior for specific database backends to ensure accurate implementation in data workflows, especially with dbplyr.

Enables task-specific AI assistants with context separation, delegation learning, and tool configuration best practices.

Ensures data assets have accurate and complete metadata, enhancing data usability and integrity for processing and delivery.

Refines prompts for GPT models (GPT-5, GPT-5.1, Codex) using OpenAI best practices to improve model output quality and task execution.

Refines prompts for Anthropic's Claude models (Opus, Sonnet, Haiku) using best practices to optimize complex task execution.