3. Data, Science & AI

Creates and optimizes AGENTS.md configuration files for AI coding agents, ensuring proper project context and agent functionality.

Builds AI agents with Vercel AI SDK, enabling tool use, tracing, and failover for LLM integrations and assistant development.

Automates quality assurance checks, metrics, and regression verification for data and image pipeline outputs, ensuring consistent quality and adherence to acceptance criteria.

Generates database migration scripts for schema changes using Drizzle ORM, supporting new tables and relationships.

Generates and edits images using Gemini AI model (text-to-image, image-to-image) via Python SDK for creative content creation.

Automates extraction and standardization of USDA WASDE report data into versioned datasets for commodity market nowcasting.

Automates financial data extraction from mining companies to calculate and explain the mining stock/metal ratio with reproducible valuation factors.

Calculates gold/silver mining gross margin proxy using metal prices and cost indicators (AISC, All-in cost), comparing current values to historical ranges.

Generates and opens an interactive backlog dashboard in the browser for data-driven project visibility.

Displays AI identity, model details, session stats, and Oracle philosophy upon user query about the AI's identity.

Forecasts relative performance between growth (Nasdaq 100) and defensive (Healthcare/XLV) sectors using US Treasury yield curve spreads.

Analyzes bank loan-deposit decoupling to track deposit trajectories and assess Federal Reserve policy transmission within banking systems.

Quantifies hedge fund positioning in agricultural futures using COT data and macro indicators to generate tradable signals.

Analyzes copper inventory trends and historical price percentiles to identify short-term price peaks and long-term value opportunities.

Detects market volatility regime shifts using 14-day smoothed ATR, providing technical levels, stop-loss, and holding period assessments for trading strategies.

Provides best practices and development patterns for AI/ML model training and application development.

Quantifies Japan's government debt burden by analyzing interest expenditure relative to tax revenue, triggered by bond yield changes, with stress testing for debt spiral risk assessment.

Analyzes US equity valuation metrics against century-long historical data to identify extreme overvaluation using percentile scoring and historical market analogies.

Analyzes demographic aging, debt dynamics, and inflation to quantify fiscal vulnerability and identify currency dilution pathways across countries.

Analyzes Japanese Government Bond (JGB) net buying/selling time series to determine record sales, consecutive selling periods, and cumulative net sales over 10+ year periods.

Applies principles from 'Designing Data-Intensive Applications' to guide system design, database selection, and distributed systems trade-offs for data-centric applications.

Analyzes public market data to evaluate correlations between interest rate volatility (MOVE), JGB yield events, and leading indicators such as VIX and credit spreads.

Analyzes daily natural gas and fertilizer price data to validate causal relationships in supply chain disruptions, identifying key turning points and lead-lag patterns.

Analyzes Federal Reserve securities' unamortized discounts for historical crisis pattern matches to validate financial stress versus accounting effects.