3. Data, Science & AI

AI-powered question answering from documents with source citations for multiple formats.

Orchestrates multi-AI pipeline execution with structured workflow guidance (plan → review → implement → review) and task file management.

Guides users through a structured multi-AI workflow for problem-solving, including planning, review, implementation, and iterative refinement.

Optimizes AI assistant configurations via prompt engineering and MCP settings to enhance performance and behavior.

Enables context-aware code implementation using the Sonnet AI model with full main context access for AI-driven applications.

Optimizes Claude's subagent usage by analyzing patterns, evaluating configurations, and creating custom agents for improved task specialization and performance.

Optimizes prompts for Claude 4.5 using best practices, guardrails, and context management to enhance AI interaction and reduce hallucinations.

Provides standardized data handling and compliance practices for Dutch government applications, ensuring adherence to national data exchange standards and quality requirements.

Analyzes SQL query execution plans and provides optimization recommendations to enhance database performance.

Enables implementation of code leveraging the Sonnet AI model with full main context access for seamless integration.

Expertly designs SQLModel schemas and manages Neon PostgreSQL databases for efficient data modeling, relationship handling, and migrations.

Designs Kafka topics, schemas, and processing logic for event-driven data pipelines in asynchronous systems.

Defines a three-layer architecture (meta-layer, Skills Layer, Application Layer) for robust orchestration of AI agents with domain capabilities and safety controls.

Executes SQL queries and data modifications on PostgreSQL databases within Docker containers, supporting SELECT, INSERT, UPDATE, and DELETE operations.

Automates selection and coordination of prompt engineering skills through domain-based classification for optimized AI interactions.

Generates long, consistent prompts for end-to-end AI tasks using the Prompt Generator (skillsmp) bridge.

Expert guidance for building AI agents and multi-agent workflows using Microsoft Agent Framework for .NET, including function tools, structured outputs, and observability.

Builds conversational AI interfaces with Chainlit, supporting LLM integrations, RAG, streaming, and multi-platform deployment for chat applications.

Builds multi-agent AI applications with OpenAI Agents SDK, supporting agent handoffs, Pydantic structured outputs, and provider-agnostic LLM workflows.

Analyzes sentiment and contrarian indicators from financial influencer discussions on PTT stock board for market trend insights.

Intelligent multi-pass PDF analyzer extracting document structure, content, and metadata via semantic inference and pdfcpu-based chunking for large documents.

Automates PostgreSQL database setup, schema/data validation, and SQL query testing with comprehensive report generation for data model verification.

Automates data modeling through step-by-step entity extraction and ER diagram generation using an immutable data model and blackboard pattern.

Provides knowledge and usage patterns for the TOON format to enable agent communication and memory persistence within the plan-marshall marketplace.