Tag: Service Deep Dive

Amazon Lex brings Alexa-grade conversational AI to your applications — supporting both voice and text across multiple channels, with generative AI capabilities powered by Amazon Bedrock. This guide covers intent-based dialog architecture, Bedrock GenAI integration (Assisted NLU, QnA Intent, BedrockAgentIntent), Visual Conversation Builder, Automated Chatbot Designer, two interaction models (request-response vs streaming), pricing, security, Amazon Connect IVR integration, and a comparison with Azure Bot Service.

Amazon Polly converts text to lifelike speech through four voice engines — Standard, Neural, Long-Form, and Generative — across 30+ languages. This guide covers all engine tiers, SSML control, custom lexicons, Speech Marks, pricing, and a comparison with Azure AI Speech.

Amazon Transcribe converts speech to text at scale — supporting 100+ languages, real-time streaming, speaker diarization, PII redaction, and specialized variants for healthcare (Transcribe Medical) and contact centers (Call Analytics). This guide covers batch and streaming modes, custom vocabularies, pricing, security, and a comparison with Azure Speech to Text.

Amazon Textract goes beyond OCR to extract structured data from invoices, forms, IDs, and loan packages using ML-powered APIs. This guide covers DetectDocumentText, AnalyzeDocument, Queries, AnalyzeExpense, AnalyzeID, Analyze Lending, Custom Queries, pricing, security, and a comparison with Azure AI Document Intelligence.

Amazon Rekognition adds enterprise-grade computer vision to your applications through simple API calls — detecting objects, faces, text, celebrities, and unsafe content without building or training ML models. This practitioner’s guide covers all core APIs, Face Liveness, Custom Labels, content moderation, pricing, responsible AI considerations, the April 2026 maintenance mode announcement, and a head-to-head comparison with Azure Computer Vision.