3. Data, Science & AI

Generates images via Google's Gemini 3 Pro Image model using structured workflow-based prompting.

Expert guide for Layer Cake, a headless Svelte framework for building responsive data visualizations with support for SVG, Canvas, HTML, and WebGL rendering.

Expertly manages AI tool workflows for Claude Code, Codex, and Gemini via CLI, enabling multi-provider prompt execution and configuration.

Enables geofencing, spatial queries, real-time tracking, and fleet visualization using PostGIS and PGRouting for location-based applications.

Expert guide for LayerChart, a Svelte-based component library enabling diverse data visualizations with unopinionated building blocks and advanced interactions.

Evaluates portfolio risk through volatility, drawdown, and key metrics for investment analysis and comparison.

Evaluates investment strategies by simulating performance on historical market data to validate returns, calculate metrics, and stress test portfolios.

Executes portfolio optimization using Mean-Variance Optimization (MVO) or Hierarchical Risk Parity (HRP) to determine optimal asset weights from financial data.

Provides prompt engineering guidance to maximize Gemini model features like multimodal processing, system instructions, and reasoning capabilities.

Provides prompt engineering guidance to optimize interactions with the Grok (xAI) model, enhancing conversational style and leveraging real-time knowledge access.

Provides prompt engineering guidance for Claude (Anthropic) to leverage XML tags, long-context, and instruction following capabilities.

Provides LLM-powered iterative search and reasoning to decompose complex queries, validate information, and generate structured analysis reports.

High-performance RAG service using Milvus for semantic search and reranking models to efficiently retrieve and rank documents from large datasets.

Enables visible reasoning steps and traceable decisions in AI execution chains for enhanced model transparency.

Provides text vectorization services converting natural language to dense vectors for semantic search, clustering, and recommendation systems.

Deep learning-based service for re-ranking document retrieval results to improve precision and top-K accuracy in search systems.

Provides high-precision text extraction from images using AI, supporting multiple languages and formats with coordinate data for document digitization and content analysis.

Constructs RyuGraph databases using code analysis results and ubiquitous language for domain knowledge modeling.

Visualizes RyuGraph database contents in Mermaid, DOT, or HTML formats for clear graph representation and analysis.

Diagnoses OpenStudio and EnergyPlus energy models, resolving geometry errors, HVAC validation, and LEED Appendix G compliance for commercial buildings.

Provides building energy modeling, code compliance verification, and performance optimization using EnergyPlus and ASHRAE standards.

Adjusts HVAC, thermal zones, and building components in OpenStudio models, runs simulations, applies measures, and manages versioned files for energy model validation.

Provides analytics for building automation systems using SkySpark, including Axon queries, trend analysis, fault detection, and performance optimization with Haystack tagging.

Analyzes EnergyPlus IDF building energy models with validation, HVAC topology, and ECM testing capabilities, supporting direct parsing and MCP tools.