Designing database schemas from scratch or optimizing existing ones is complex. This skill provides proven patterns for normalization, indexing, migrations, and performance optimization across SQL and NoSQL databases.

Financial modeling requires complex calculations and scenario testing. This skill automates DCF valuations, sensitivity analysis, and Monte Carlo simulations for accurate investment decisions.

This skill calculates key financial ratios from balance sheets, income statements, and cash flow data. It provides investment analysis with industry benchmarks and actionable recommendations.

Marketing teams need to quantify campaign effectiveness across multiple platforms. This skill calculates engagement rates, ROI, and cost efficiency metrics to deliver data-backed optimization recommendations.

Building ML pipelines requires expertise in model deployment, monitoring, and MLOps practices. This skill provides production-ready tools for deploying ML models, integrating LLMs, and building scalable RAG systems with enterprise-grade reliability.

Generic prompts produce inconsistent AI responses. This skill provides proven patterns and optimization tools for production-grade LLM interactions. Transform vague requests into precise, reliable outputs using chain-of-thought, few-shot learning, and structured prompting techniques.

Design rigorous A/B tests and experiments with proper statistical power analysis. Build production-ready predictive models using proven feature engineering techniques and machine learning frameworks.

Create enterprise-grade computer vision solutions with object detection, image segmentation, and real-time video analysis. This skill provides expert guidance on PyTorch, OpenCV, YOLO, and vision transformers for production deployment.

Duplicate images waste storage space and bias machine learning models. This skill uses deep learning embeddings to identify and remove exact duplicates and near-duplicate images from FiftyOne datasets. It automates the detection process and helps you clean datasets before training.

Understanding complex image datasets requires seeing how samples relate in embedding space. This skill guides you through computing embeddings and using UMAP or t-SNE to create 2D visualizations that reveal clusters, outliers, and class distributions in your FiftyOne datasets.

Users struggle to extend the FiftyOne computer vision platform with custom functionality. This skill provides step-by-step guidance for building operators and panels that integrate seamlessly with the FiftyOne App.

Loading local media files into FiftyOne and running ML model inference requires understanding dataset creation, label imports, and model application. This skill provides step-by-step guidance for creating datasets, importing labels in standard formats, and applying zoo models for detection, classification, and segmentation tasks.

Dataset import often requires manual format detection and configuration. This skill auto-detects media types, label formats, and sensor groupings to create properly structured FiftyOne datasets.

Database management requires deep knowledge of SQL patterns and Supabase features. This skill provides expert guidance for creating tables, writing migrations, configuring RLS policies, and troubleshooting database issues with confidence.

Working with storytelling platforms requires consistent patterns for extracting themes, quotes, and AI insights. This skill provides ready-to-use patterns for Supabase queries, theme matching, and AI analysis integration.

Building applications with SurrealDB requires understanding its multi-model capabilities including graph relationships, vector embeddings, and real-time subscriptions. This skill provides comprehensive guidance for using the Python SDK to implement CRUD operations, semantic search, and graph traversal patterns.

Building data extraction pipelines with LLMs requires consistent, validated JSON outputs. This skill provides expert guidance on using Cohere v2 JSON Schema mode for reliable entity extraction, classification, and structured data pipelines.

Extract detailed text and visual data from PDF pages. This skill captures font metadata, text positions, and rendered images to enable accurate AI-driven HTML generation workflows.

Users struggle to connect document repositories with AI models for accurate, context-aware responses. This skill provides a complete RAG pipeline including document ingestion, vector search with Qdrant, and personalized prompt generation for specialized knowledge bases.

Database operations are complex and error-prone when done manually. This skill provides guidance on using Drizzle ORM with NeonDB Postgres for type-safe database queries, schema management, and migrations.