3. Data, Science & AI

Provides core functionality for creating static plots, layouts, and modular figures using the Makie visualization library in Julia.

Generates and manages database schema migrations using Kysely, enabling table creation, column additions, and schema modifications.

Generates dynamic dashboards, animations, and interactive visualizations using Makie with Observables and UI widgets.

Optimizes data formats (JSON/YAML), validates schemas, and implements high-performance serialization for efficient LLM data transmission and application integration.

Splits documents into manageable chunks for efficient AI processing, particularly in RAG pipelines.

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

Retrieves detailed technical properties and metadata from image files including resolution, channels, transparency, and compression type.

Provides foundational principles for AI-powered development workflows including agent management, token optimization, and modular architecture patterns.

Orchestrates implementation phase by spawning AI agents for coding (implementer) and code review (reviewer) tasks.

Modern asynchronous ORM providing type-safe database models and optimized query execution for efficient data handling.

Enterprise context management for LLM applications with token budget optimization and persistent state handling across sessions.

Intelligently summarizes context to prevent LLM context window overflow, triggered automatically or manually.

Verifies mathematical operations, financial calculations, and data transformations for accuracy in data science and AI workflows.

Manages version-controlled database schema migrations for SQLAlchemy projects, ensuring consistent database evolution.

Dispatches independent AI agents to process non-dependent tasks concurrently without shared state or sequential constraints.

Configures Markov Decision Process (MDP) elements including observations, rewards, and termination conditions for reinforcement learning environments.

Provides secure connection to Vayyar's PostgreSQL databases via Google Cloud SQL Proxy for querying, debugging, and data analysis in US/EU regions.

Executes Siril job files for astrophotography processing, handling calibration, stacking, registration, and composition of LRGB/SHO/HOO images.

Optimizes long-form conversation accuracy through a four-layer memory architecture and dynamic context pruning for enhanced AI dialogue systems.

Enhances domain intelligence through an AI-driven protocol for analyzing and optimizing domain-related data and strategies.

Fetches real-time and historical market data from multiple sources including Yahoo Finance, FRED, and Stooq for stock, index, and economic indicator analysis.

Extracts and analyzes text, images, tables, and entities from PDF documents using NLP and data processing techniques.

Compares data model designs to select optimal schemas for new data modeling or schema refactoring tasks.

Provides seamless integration with OpenAI's LLMs (GPT-4o, o1, etc.) for text generation, reasoning, and code assistance via Python SDK.