Skill UI

Client library to manage Azure Machine Learning resources—workspaces, datasets, compute, jobs, models, environments, and pipelines—directly from Python with MLClient authentication, configuration, and best practices for reproducible workflows.

Client library to manage Azure Machine Learning resources—workspaces, datasets, compute, jobs, models, environments, and pipelines—directly from Python with MLClient authentication, configuration, and best practices for reproducible workflows.

Guides how to stretch limited LLM context windows by compressing, masking, caching, and partitioning information so agents process longer conversations or documents without enlarging models; includes when to trigger optimizations and concrete patterns for summaries, masking, KV-cache reuse, and budget tracking.

A specialized skill for optimizing LLM applications handling long conversations and large datasets. This guide teaches advanced context engineering techniques, addressing issues like context rot, token limits, and the lost-in-the-middle problem. Capabilities include intelligent summarization, context trimming, information prioritization, and strategic context routing to ensure critical information is retained and utilized effectively throughout the dialogue.

Provides production-tested patterns for building AI agents that autonomously discover, evaluate, and acquire token listings for cryptocurrency exchanges, covering multi-source intelligence, 100-point scoring, wallet forensics, outreach pipelines, micropayment orchestration, and ERC-8004 registration.

Guides teams through audio preprocessing, connection pooling, model selection, streaming workflows, concurrency controls, and caching to accelerate Deepgram transcription and cut latency across large-scale pipelines.

Deploying Machine Learning Models automates serving trained models to production by analyzing requirements, generating API and infrastructure code, and orchestrating deployment workflows with validation, error handling, and monitoring baked in.

Feature Engineering Toolkit automates creating, selecting, and transforming dataset features to improve machine learning models; use it whenever you need new interactions, automatic selection, scaling, or encoding to boost accuracy and interpretability.

Guides teams through choosing and implementing validated Exa deployment architectures—from direct search to cached semantic layers and RAG pipelines—so integrations scale from simple apps to high-query, AI-enhanced solutions.

Guides teams through reducing Exa AI search API bills by adding query caching, picking cost-effective search types, limiting result counts, deduplicating batch queries, and monitoring budget and usage alerts.

Establish circuit breakers, caching, retries, fallback chains, and quality monitoring for Exa neural search integrations so production services stay resilient under rate limits, empty results, or slow responses.

Provides interpretable explanations for ML predictions using SHAP, LIME, and other techniques, helping debug model behavior, highlight feature importance, and communicate fairness insights to stakeholders.