Skill UI

Designs and implements production-grade RAG systems with document chunking, embeddings, vector stores, hybrid search, reranking, and evaluation, ideal for knowledge-grounded AI apps that need semantic retrieval, similarity search, and context augmentation.

Auto-activating assistant that guides AWS Sagemaker endpoint deployment, from best practices to production-ready code, configurations, and validation for inference serving and monitoring workflows.

Comprehensive skill covering the full lifecycle of visual AI systems. Expertise includes object detection (YOLO, Faster R-CNN), instance and semantic segmentation (SAM, Mask R-CNN), and video analysis. The workflow guides users through dataset preparation, selecting optimal architectures, training using frameworks like PyTorch, and achieving production readiness via model optimization (ONNX, TensorRT) and deployment on various edge or cloud targets.

Sentence Transformers provides 5000+ pre-trained models that generate sentence, text, and image embeddings for RAG, semantic search, clustering, and multilingual production workloads without relying on external APIs.

SGLang is a high-performance serving framework for LLMs and VLMs that uses RadixAttention prefix caching to deliver structured outputs (JSON/regex/grammar) and agentic workflows with function calls, providing up to 5× faster inference than vLLM across multi-GPU production deployments.

This pattern provides established best practices for implementing efficient similarity search in production systems. It is crucial for building semantic search engines, implementing Retrieval-Augmented Generation (RAG), creating advanced recommendation systems, and optimizing retrieval performance when scaling to millions of vectors.

Guides post-training of GLM-class LLMs with Megatron-LM and SGLang, focusing on GRPO rollout generation, custom data buffers, and tight RL scaling workflows for research and production use.

Provides production-ready TypeScript and Python patterns for Speak’s REST API, covering pronunciation assessment, conversation practice, batch handling, and rate-limited workflows so teams can integrate language-learning features with consistent SDK usage.

Automates guidance for streaming inference setup, covering best practices in ML deployment, serving, monitoring, and production-ready configurations whenever you ask about streaming inference patterns or implementation.

Assists with tensorboard visualizer tasks for ML training by guiding setup, best practices, and experiment tracking whenever tensorboard visualization is mentioned, producing production-ready configs and validated outputs.

Automates TensorFlow SavedModel deployment support with step-by-step guidance, best-practice patterns, production-ready code/config generation, and validation for serving, monitoring, and optimization scenarios.

Optimizes large language model inference on NVIDIA GPUs with TensorRT, delivering 10-100× faster throughput, quantized precision (FP8/INT4), multi-GPU scaling, and serving-ready tooling for production deployments.