3. Data, Science & AI

Advises on session reuse or new session creation for multi-model AI interactions before calling models like Codex or Gemini.

Provides strategies to minimize token usage in LLM applications, significantly reducing costs by 60-80%.

Assists in setting up, optimizing, and auditing Google Analytics 4 and tag manager implementations for accurate data tracking and measurement.

Performs data exploration, statistical analysis, and creates visualizations for data-driven insights.

Enables unified access to multiple LLM models via OpenRouter, comparing responses, tracking costs, and optimizing model selection for AI tasks.

Tracks token usage and costs for AI services with customizable thresholds to manage spending and avoid pricing overages.

Generates visual content by converting text prompts into images using artificial intelligence models.

Tracks large cryptocurrency holder transactions and institutional movements on blockchain networks for market analysis and insight.

Builds evaluation frameworks to test AI agent performance, validate context engineering, and measure improvements over time.

Analyzes code repositories for AI agent development readiness, scoring key aspects with ASCII dashboards to optimize for AI-assisted development.

Manages WOPR AI session lifecycle including creation, inspection, and context handling via CLI or API.

Coordinates data pipelines, quality validation, and ML infrastructure for AI-driven crypto trading strategies.

Expert guide for crafting effective prompts, optimizing LLM interactions, and debugging agent behavior using proven techniques.

Controls Google Gemini AI via MQTT WebSocket, enabling message sending and browser tab management through voice commands.

Extracts structured data from web pages, parses PDFs, and crawls websites using Firecrawl API for data processing workflows.

Configures Model Context Protocol (MCP) servers for IDEs, enabling AI context management and troubleshooting in development environments.

Detects missing requirements in user input and generates targeted clarification questions to ensure complete system understanding.

Queries schedule.md using DuckDB's markdown extension to retrieve event information for user requests about schedules and calendars.

Enables embeddings-based semantic search for similar content discovery, research papers, and category-specific information via Exa API.

Enables self-referential AI loops with session isolation for iterative refinement in LLM agent workflows.

Crafts optimized prompts for AI models to generate high-performing social media content and marketing copy.

Generates type-safe AI content with guaranteed structure using Gemini and Zod validation for seamless application integration.

Performs radio astronomy data calibration, flagging, imaging, and mosaicking for science-quality image generation.

Manages data ingestion, staging, metadata, and storage layout for efficient pipeline operations, with specialized support for HDF5 file lifecycle and product I/O.