Azure AI Search

Information Retrieval at Scale:
- Azure AI Search provides secure information retrieval over user-owned content in both traditional and conversational search applications.
- It’s foundational for any application that surfaces text and vectors.
Capabilities:
- Search Engine: Azure AI Search offers a powerful search engine for:
  - Vector Search: Efficiently searching vectors (e.g., embeddings) associated with data.
  - Full Text Search: Searching through textual content.
  - Hybrid Search: Combining vector and text search.
- Rich Indexing:
  - Data Chunking and Vectorization (Preview): Efficiently indexing large data chunks and vectors.
  - Lexical Analysis: Processing text for indexing.
  - Optional AI Enrichment: Enhancing content extraction and transformation using AI.
- Rich Query Syntax:
  - Supports various query types, including vector queries, fuzzy search, autocomplete, and geo-search.
Architectural Overview:
- A search service sits between:
  - External Data Stores: Contain un-indexed data.
  - Client Apps: Send query requests to a search index and handle responses.
- In your client app, you define the search experience using Azure AI Search APIs, including relevance tuning, semantic ranking, and more.
Integration with Azure Services:
- Indexers: Automate data ingestion from Azure data sources.
- Skillsets: Incorporate AI from Azure services (e.g., image processing, NLP) or custom AI created in Azure Machine Learning or Azure Functions.
Indexing and Querying:
- Indexing: Loads content into the search service, making it searchable.
  - Inbound text is processed into tokens and stored in inverted indexes.
  - Inbound vectors are stored in vector indexes.
  - JSON format is used for indexing.
- Querying: Happens once an index is populated.
  - Client apps send query requests to the search service.

Azure AI Translator

Azure AI Translator is a cloud-based service that uses AI to translate text and documents between languages in near real-time. It supports translation in over 90 languages and dialects and can be used with any operating system. The service also offers customizable translation models that can better understand industry-specific terminology or pronouns This makes it a powerful tool for technical people who need to work with multilingual content or communicate with colleagues and customers in different languages.

Azure AI Language

Information Retrieval at Scale:
- Azure AI Language offers a powerful search engine for:
  - Vector Search: Efficiently searching vectors (e.g., embeddings) associated with data.
  - Full Text Search: Searching through textual content.
  - Hybrid Search: Combining vector and text search.
Rich Indexing and Querying:
- Indexing: Loads content into the search service, making it searchable.
  - Inbound text is processed into tokens and stored in inverted indexes.
  - Inbound vectors are stored in vector indexes.
  - JSON format is used for indexing.
- Querying: Happens once an index is populated.
  - Client apps send query requests to the search service.
Capabilities:
- Extract Information:
  - Use Natural Language Understanding (NLU) to extract information from unstructured text.
  - Identify key phrases, Personally Identifiable Information (PII), and named entities.
  - Summarize text-based content.
- Classify Text:
  - Detect language or classify sentiment.
  - Customize classification models over your dataset.
- Answer Questions:
  - Provide answers to questions in unstructured texts.
  - Create custom question-and-answer pairs.
- Conversational Language Understanding:
  - Classify conversational utterances and extract detailed information.
- Translate Text:
  - Use cloud-based neural machine translation for multi-language solutions.
Integration with Azure Services:
- Indexers: Automate data ingestion from Azure data sources.
- Skillsets: Incorporate AI from Azure services or custom AI.