Today, a significant announcement has been made regarding Azure AI Search, aimed at supporting customers in developing production-ready generative AI applications. The enhancements include a substantial increase in storage capacity and vector index size within Azure AI Search, provided at no extra cost. This update enables customers to seamlessly execute retrieval augmented generation (RAG) operations at any scale without compromising on cost efficiency or performance.
This article will guide you through the following aspects:
- How customers can achieve enhanced scalability cost-effectively with the latest modifications.
- The reliability of Azure AI Search in managing large RAG workloads.
- The application of advanced search methodologies to navigate intricate data, fostering innovation beyond conventional boundaries.
Azure AI Search
Facilitating precise, personalized responses in generative AI applications
Introduction of Enhanced Scalability and Performance at a Reduced Cost with Azure AI Search
Azure AI Search has significantly elevated its vector and storage capacity, offering customers increased scalability, superior performance, and more value for their data investment.
The capacity and computational resources available for Azure AI Search’s Basic and Standard tiers have been augmented in specific regions.
Here’s a breakdown of the key improvements:
- Up to an 11x increase in vector index size.
- A 6x boost in total storage.
- Twofold enhancement in indexing and query throughput.
These enhancements empower customers to deliver top-notch user experiences across various scales. With the ability to scale generative AI applications to a multi-billion vector index within a single search instance, customers can ensure optimal speed and performance.
Supporting Extensive RAG-Based Applications with a Reliable Enterprise Retrieval System
Over half of the Fortune 500 companies rely on Azure AI Search to handle their crucial enterprise search and generative AI tasks. Noteworthy entities such as OpenAI, Otto Group, KPMG, and PETRONAS utilize Azure AI Search to manage their retrieval augmented generation (RAG) workloads.
When OpenAI unveiled their RAG-powered “GPTs” and the Assistant API during OpenAI DevDay 2023, they required a robust retrieval system capable of meeting unprecedented demand and scalability. Azure AI Search emerged as the solution of choice for OpenAI, given its capacity to support large-scale RAG operations on the internet.
Azure AI Search now caters to RAG functionalities for ChatGPT, GPTs, and the Assistant API, offering search capabilities to products like the GPT Store. Whether it’s a search query or file addition within these products, Azure AI Search serves as the backbone of the retrieval system.
As of November 2023, ChatGPT alone attracts 100 million weekly visitors, with over 2 million developers integrating its API. The introduction of custom GPTs led to the creation of 3 million GPTs within two months, reflecting a global user base engaging in RAG activities at an extensive scale.
Building Enhanced Applications with a Modern, Advanced Retrieval System
Professionals in sectors such as healthcare, telecommunications, and professional services acknowledge that relying solely on vector search is insufficient for developing a functional generative AI application. Different retrieval techniques are better suited for distinct use cases, necessitating a blend of approaches for optimal performance.
Azure AI Search equips applications with a range of out-of-the-box strategies, including hybrid retrieval and semantic reranking, enabling developers to achieve their objectives efficiently.
Case Study: Telus Health Leveraging Advanced RAG for Customer Care
Telus Health, based in Canada, is renowned for its tech-driven solutions catering to employers, individuals, healthcare practitioners, and insurers. The organization introduced a customer support platform to address user queries regarding health plans and offer website navigation guidance. Initially relying solely on vector search proved inadequate, prompting Telus Health to turn to Azure AI Search for its diverse search technologies.
The Guide Team at Telus Health played a pivotal role in refining the search strategy and leveraging AI Search effectively to enhance the platform’s functionality. By adopting hybrid search with semantic reranking, Telus Health successfully addressed inquiries related to client-specific documents and website navigation, significantly enhancing the platform’s accuracy and responsiveness.
Empowering Brands with Multi-Vector Retrieval: NIQ Brandbank’s Innovative Solution
NIQ Brandbank assists Fast-Moving Consumer Goods (FMCG) brands in optimizing their online presence by providing rich, relevant content and imagery for their digital shelf, enabling them to stand out in the market.
NIQ Brandbank’s Content Health+ solution empowers brands with actionable insights to enhance their online visibility and product placement across retailer search results, ultimately boosting sales.
By employing a user-friendly interface, Content Health+ leverages multi-vector search to offer tailored recommendations based on research findings by the NIQ Data Impact team. This approach enhances product performance on the digital shelf by identifying key attributes influencing organic placement and providing valuable recommendations for brand enhancement.
To ensure optimal performance, Content Health+ integrates hybrid multi-vector search and semantic ranking, allowing for a holistic approach to data retrieval and recommendation systems.
In the words of Gabriel Harris PhD, Principal Data Scientist:
“Azure AI Search enables us to implement hybrid multi-vector search, combining text and image embeddings with semantic ranking to prioritize the most contextually relevant products. From seamless data ingestion to integration with Azure Machine Learning, Azure AI Search has been instrumental in bringing the Content Health+ application to life.”
Explore the Capabilities of Azure AI Search
With these recent enhancements, Azure AI Search empowers AI systems to retrieve information efficiently at scale. Customers can embark on innovative ventures with confidence, leveraging cutting-edge retrieval technology within Azure AI Search for a robust enterprise-ready foundation.
