Skill UI

A comprehensive guide to optimizing Groq API usage and managing inference costs. Learn techniques such as smart model routing (selecting smaller models for simple tasks), minimizing prompt tokens, batching requests to reduce overhead, and implementing caching for deterministic calls. This guide helps developers achieve significant cost savings and improve system efficiency when handling high-volume AI workloads.

Comprehensive guide for deploying applications powered by the Groq inference API. Covers best practices and specific setup instructions for multiple production environments, including Vercel Edge Functions, Google Cloud Run, Docker containers, and other containerized platforms. Essential for setting up robust, low-latency, production-ready AI services.

This guide provides comprehensive instructions for installing and configuring the official Groq SDK in both Python and TypeScript environments. It covers dependency setup, secure API key management using environment variables, and provides code examples to verify a successful connection to Groq's ultra-fast LLM inference API. Essential for initiating any project powered by Groq.

A comprehensive guide to implementing advanced asynchronous patterns with Mistral AI. This skill covers using the Agents API for stateful, multi-turn workflows, the Batch API for cost-effective bulk inference, and event-driven streaming architectures. It is essential for building robust, long-running, or high-volume AI applications that require reliable background job handling.

This skill automates the setup and configuration of Ollama, allowing users to deploy Large Language Models (LLMs) locally. It performs system assessment (OS, GPU, RAM), selects optimal models based on hardware availability, and handles installation across macOS, Linux, and Docker. Use this when you require free, self-hosted AI alternatives, need offline inference capabilities, or wish to eliminate reliance on costly hosted API services.

Master the Zod library for defining robust, type-safe schemas in TypeScript. This expert covers advanced validation techniques, including primitives, objects, unions, and discriminated unions. It specializes in type inference (`z.infer`), data coercion, complex refinements, and seamless integration with modern frameworks like React Hook Form, Next.js, and tRPC, ensuring consistency across your application state and API inputs.

Modal is a serverless cloud platform for running Python code on demand, including GPU-accelerated AI/ML workloads. Use for deploying models, serving web APIs, batch processing, and scaling Python applications to cloud containers with automatic scaling from zero to thousands of instances. Supports custom container images, persistent storage, and sub-second cold starts for low-latency inference.

Automate the entire deployment lifecycle for ML inference services on CoreWeave GPU cloud. This workflow integrates unit testing, container image building (to GHCR), and rolling deployment to CoreWeave Kubernetes clusters via GitHub Actions. Ideal for ensuring stable, high-performance deployment of GPU-dependent models. Key features include validating GPU manifest requests and managing deployments using standard Kubernetes APIs.

A comprehensive reference guide for common errors encountered when interacting with the Together AI API. This includes error codes and solutions for inference, fine-tuning, and model deployment, covering authentication, rate limits, validation failures, and billing issues. Essential for developers integrating OpenAI-compatible LLM services.

A comprehensive workflow guide for fine-tuning open-source Large Language Models (LLMs) using the Together AI API. This process covers preparing structured training data (JSONL format), uploading the file, initiating the fine-tuning job, monitoring the training status, and finally deploying the custom model for inference. Ideal for developers needing to customize models for specific enterprise tasks, such as customer support or specialized content generation.

This guide covers essential security best practices for utilizing Together AI's OpenAI-compatible API for inference, fine-tuning, and model deployment. It provides code examples and architectural advice on secure API key management, robust webhook signature verification, input validation using schemas (like Zod), and implementing data redaction pipelines to protect sensitive information in logs and datasets. Essential for building secure AI applications.

A critical discipline for managing expenses when integrating paid AI or inference APIs. This guard treats dollar cost as a third complexity dimension, mandating the definition of cost contracts (max calls per run, max budget per run, max budget per day). It enforces best practices for bounding loops, retries, and fan-out, ensuring that code bugs cannot lead to unexpected overspending. Best practice requires configuring hard caps both in the code and at the provider dashboard for defense in depth.