Skip to content

11-ai-services.md

Latest commit

 

History

History
101 lines (79 loc) · 3.69 KB

11-ai-services.md

File metadata and controls

101 lines (79 loc) · 3.69 KB

AWS AI Services

Amazon Lex

  • Build conversation bots
  • 💡 Use-cases:
    • Customer service chatbots
    • Voice assistants
    • FAQ automation
    • Order processing bots
    • Appointment scheduling

Amazon Polly

  • Text to speech.
  • Uses deep learning to synthesize natural-sounding human speech
  • Provides life-like voices across different languages
  • 💡 Use-cases:
    • Build speech-enabled apps
    • Accessibility features
    • Audiobooks
    • Voice announcements
    • IVR systems
    • E-learning content

Amazon Comprehend

  • 📝 Derive and understand valuable insights from text within documents.
  • Uses NLP to extract insights about the document contents without any preprocessing.
  • Processes text files to identify entities, key phrases, language, sentiments.
  • 💡 Use-cases:
    • searching social networking feeds for mentions of products
    • scanning entire document repository for key phrases
    • determining topics in a set of documents
    • PII detection
    • sentiment analysis
    • document classification
    • entity extraction
    • content moderation

Amazon Rekognition

  • 📝 Automates image recognition and video analysis for your applications
  • Features: real face identification, face detection and analysis, text detection
  • Can detect custom objects (e.g. brand logos) using AutoML (10 image examples enough)
  • 💡 Use-cases: Content moderation, security surveillance, photo tagging, identity verification, brand monitoring

Amazon Textract

  • 📝 Automatically extract text, handwriting, layout elements and data from scanned documents.
  • Beyond simple OCR: identify, understand and extract specific data from documents.
  • Features: forms extraction, table extraction, handwriting recognition, layout analysis
  • Supports various formats: PDFs, images (JPEG, PNG, TIFF)
  • 💡 Use-cases:
    • Automating loan processing
    • Extracting information from invoices and receipts
    • Document digitization
    • Form automation
    • Insurance claims processing
    • Financial document analysis

Amazon Translate

  • Neural machine translation service
  • 💡 Use-cases: localize content, analyze large volumes of text for cross-lingual communication

Amazon Bedrock

  • Serverless service to build and scale generative AI applications with foundation models.
  • Foundation Model
    • Model performing wide variety of general tasks
    • E.g. understanding language, generating text and images, and conversing in natural language
    • Large deep learning neural networks.
    • Rather than developing AI from scratch, allows data scientists to use it as starting point to develop ML models.

Amazon Kendra

  • Enterprise search powered by machine learning
  • Lets you ask questions in neutral language or in keywords
  • Can search in S3, RDS, Microsoft SharePoint, ServiceNow and Salesforce.

Amazon SageMaker

  • Fully managed machine learning service
  • Allows building, training and deploying ML models.
  • Provides integrated Jupyther notebook instances for exploration and analysis.
  • Provides common ML algorithms optimized for large data in a distributed environment.
  • Native support for bring-your-own algorithms and framework.

Amazon Forecast

  • Fully managed machine learning service for time series forecasting
  • Uses algorithms like ARIMA, ETS, DeepAR+, and Prophet to generate accurate predictions

Amazon Fraud Detector

  • Fully managed fraud detection service using machine learning
  • Pre-built fraud detection models based on 20+ years of Amazon's fraud detection experience
  • Custom rule creation with real-time scoring APIs
  • Model interpretability features to understand prediction drivers
  • Pay-per-prediction pricing model