3. Data, Science & AI

Generates a RAG pipeline script to crawl textbook content from sitemap, chunk, embed with Gemini, and store in Qdrant per MCP documentation.

Provides a chat endpoint for agent-based RAG reasoning with input validation and structured response formatting.

Assists in designing Neural Quantum States architectures, integrating classical samplers with qiskit-addon-sqd, and analyzing sample-efficiency in quantum chemistry.

Generates JSON schemas and Python handlers for Qdrant-based retrieval tools in RAG agents to search sitemap-crawled textbook content with URL citations.

Generates Hugging Face Spaces configuration files with runtime setup, environment variables, and deployment instructions for AI model hosting.

Creates reusable embedding functions using Gemini models via LangChain with error handling and batching for sitemap-crawled content.

Provides SQL query patterns, optimization techniques, and database best practices for PostgreSQL, MySQL, and SQLite.

Provides AI-powered financial management for PocketSmith, including transaction categorization, tax compliance (ATO), scenario analysis, and budget optimization via API integration.

Creates YAML configuration files for migrating data, schemas, and ACLs between Kafka-compatible clusters including Redpanda, Confluent Cloud, and AWS MSK.

Generates and manages naming conventions for data entities, AI metadata, and labeling in machine learning workflows.

Converts dates to YYYY-MM format for standardization, extraction, and monthly data aggregation in reporting and analysis.

Establishes best practices for building Mastra agents, including tool creation and workflow orchestration in AI systems.

Automates translation, PDF processing, and code generation via Gemini CLI.

Assists in creating business analytics reports, dashboards, and visualizations using Recharts and D3.js for metrics like member growth and revenue.

Converts human-readable dates to ISO 8601 format (YYYY-MM-DD) for data standardization and machine processing.

Provides persistent memory for AI agents with semantic, episodic, and procedural types to enable context retention, history tracking, and personalized learning across sessions.

Orchestrates autonomous agent teams for complex task completion through dynamic agent selection, principle enforcement, and persistent sessions.

Interactive HTML viewer for visualizing and verifying parsed spatial omics metadata against source documents.

Provides foundational principles for AI-powered development workflows, including SPEC-First TDD, token optimization, and modular architecture.

Creates realistic synthetic CSV data with configurable headers, data types, and row counts for testing, development, and demonstration purposes.

Enables comprehensive cardiac development analysis through spatial multi-omics data integration, providing cellular and molecular insights.

Manages enterprise AI sessions with optimized token usage and persistent state for efficient context handling in LLM applications.

Specializes in creating token-efficient agent and workflow definitions using YAML, inspired by BMAD Method patterns for AI applications.

Analyzes text using Utterance Emotion Dynamics (UED) to reconstruct emotional states, trajectories, and patterns with psychological insight.