3. Data, Science & AI

Retrieves specific vector entries by key from Agentuity's cloud platform, requiring authentication for secure data access.

Provides vector storage statistics for AI/ML data management on Agentuity cloud platform, enabling performance monitoring of embedding databases.

Command-line tool for semantic similarity vector searches on the Agentuity cloud platform, requiring authentication for secure data retrieval.

Executes authenticated SQL queries on Agentuity cloud database for platform data management and operations.

Manages vector storage operations on Agentuity cloud platform, enabling addition or update of AI embeddings with authentication.

Enables implementation of data validation rules, quality metrics, and cleansing strategies to ensure data accuracy and reliability.

Designs ETL patterns, data pipeline orchestration, and incremental loading strategies for efficient data processing systems.

Designs efficient database schemas, optimizes SQL queries with proper indexing, and manages data operations according to best practices.

Designs dimensional models, star/snowflake schemas, and data warehouse structures to optimize analytics and reporting workflows.

Comprehensive guide for managing AI model selection, context, and multi-model collaboration via The Force MCP server interface.

Analyzes large codebases using recursive language model patterns to handle contexts exceeding 50K tokens and enable deep multi-file analysis.

Automates invoice data extraction from PDFs and images to Excel using AI-powered vision recognition.

Visualizes dataset structure in 2D using UMAP or t-SNE embeddings to explore clusters, outliers, and data distribution.

Evaluates AI model predictions against ground truth using COCO, Open Images, or custom protocols for metrics like mAP and confusion matrices in detection, classification, and segmentation tasks.

Exports FiftyOne datasets to standard formats (COCO, YOLO, VOC, CVAT, CSV) for AI model training and data sharing.

Imports datasets into FiftyOne with automatic format detection for images, videos, point clouds, and label formats like COCO and YOLO.

Automates duplicate image detection in datasets using deep learning-based similarity computation for efficient deduplication.

Runs ML model inference (detection, classification, segmentation) on FiftyOne datasets for AI workflows including data import and prediction tasks.

Manages Databricks Unity Catalog resources for data governance, including catalogs, schemas, and tables with proper naming and updates.

Executes Python/SQL code on Databricks clusters for stateless validation and stateful iterative debugging of data transformations and pipelines.

Builds end-to-end ML pipelines on Databricks: data exploration, feature engineering, model training, MLflow tracking, and Unity Catalog deployment.

Builds production data engineering pipelines on Databricks using medallion architecture (Bronze/Silver/Gold) for ETL, transformation, and quality checks.

Executes ML model benchmarking and evaluation, measuring performance metrics and inference speed for model optimization.

Deploys and manages NVIDIA Triton Inference Server for efficient AI model serving and inference operations.