Skill UI

A systematic workflow designed to significantly improve the reliability and performance of existing AI agents. It covers detailed performance analysis, including gathering metrics (e.g., task success rate, latency), classifying failure modes, and advanced prompt engineering techniques (like Chain-of-Thought and Constitutional AI). The process culminates in structured A/B testing and controlled deployment, ensuring measurable, robust improvements.

The Azure AI VoiceLive SDK for Java enables real-time, bidirectional voice conversations with AI assistants. Using WebSocket technology, this SDK supports low-latency streaming, advanced features like turn detection, noise reduction, and customizable voice selection (OpenAI/Azure). It is ideal for developers building sophisticated voice bots or interactive voice applications in Java.

A Python implementation utilizing the Azure Speech-to-Text REST API for transcribing short audio files (up to 60 seconds). This method avoids the need for the dedicated Speech SDK, allowing developers to perform speech recognition simply by making HTTP POST requests. It supports both simple and detailed response formats and recommends chunked transfer for lower latency.

This skill details advanced techniques for enhancing the effective context capacity of LLMs. It outlines four core strategies—Compaction, Observation Masking, KV-Cache Optimization, and Context Partitioning—to prevent context limits from constraining task complexity. By strategically managing token usage, users can reduce operational costs, minimize latency, and implement robust, long-running agent systems for handling large documents or conversations.

This guide demonstrates how to implement robust, real-time, low-latency streaming transcription using Deepgram's WebSocket API. It covers advanced audio handling, including microphone input capture (via Sox), speaker diarization, utterance boundary detection, and implementing automatic reconnection logic. Essential for building professional voice AI applications, live captioning services, or voice interfaces.

A comprehensive guide to using Groq's high-speed LPU for chat completion. This resource provides code examples in TypeScript and Python, covering basic chat interactions, streaming responses, and advanced multimodal features. Ideal for developers building high-performance, low-latency AI applications using the Groq API.

This guide details a production-ready reference architecture for applications built on Groq's LPU inference API. It covers sophisticated model routing based on latency, quality, or cost needs, implementing streaming pipelines, and establishing robust multi-provider fallback chains (e.g., Groq primary to OpenAI backup). Ideal for designing and reviewing complex, resilient LLM systems.

Automates inference latency profiler tasks in ML deployment scenarios, offering step-by-step guidance on model serving, MLOps pipelines, monitoring, and production optimization, generating production-ready code and validating outputs against best practices.

A comprehensive guide to setting up production-grade observability for LangChain applications. It integrates LangSmith tracing for zero-code debugging, custom callbacks for detailed metric collection (latency, token usage, error rate), and Prometheus exporters for robust alerting. This allows developers to monitor LLM pipelines, track performance metrics, and manage operational health in real-time.

Track and improve production LLMs with Langfuse observability: trace API calls, manage prompts, score evaluations, integrate with LangChain/LlamaIndex/OpenAI, and keep dashboards on cost, latency, and quality.

Implement comprehensive end-to-end tracing for Large Language Model (LLM) calls, chains, and complex agents using Langfuse. This workflow demonstrates techniques like OpenAI drop-in wrappers, manual span creation for RAG pipelines, tracking streamed responses, and integrating with various LLM providers (OpenAI, Anthropic). Use this to monitor latency, track token usage, and debug complex AI features in production.

LangSmith centralizes tracing, evaluation, and monitoring for LLM applications, letting teams debug prompts, run regression tests, collect latency/cost metrics, and integrate with OpenAI, Anthropic, LangChain, or LlamaIndex to keep production systems observable.