Leveraging Azure AI services (AI Search) for different data source.

Question

Hello,

We are working on a Proof of Concept (PoC) to leverage Azure AI services (AI Search), enabling users of our web application to query data using natural language. Our use case involves scenarios where the data source could be a database hosted on any cloud platform, such as Azure or AWS,On-Premise or an API developed to fetch data from ERP systems like SAP or JD Edwards. We would like to know if this approach is feasible and what Azure tools and licensing would be required to develop this PoC. Thanks

Answer 1

Hello Avinash Vyas,

Welcome to the Microsoft Q&A and thank you for posting your questions here.

I understand that you would like to leverage Azure AI services (AI Search) for different data source.

Your plan to incorporate Azure AI services for natural language queries in your web application is viable. Below are the key Azure tools and services that can help you develop a solid Proof of Concept (PoC). Recommended Azure Services are the followings:

1. Azure Cognitive Search provides advanced search functionality with built-in natural language processing (NLP) for interpreting user queries. The key features support full-text search, faceted navigation, filtering, and semantic search - https://techcommunity.microsoft.com/blog/azure-ai-services-blog/best-practices-for-using-azure-ai-search-for-natural-language-to-sql-generation-/4281347
2. Azure AI Language enhances NLP capabilities, including sentiment analysis, key phrase extraction, and named entity recognition. it helps refine user queries for better search accuracy - https://learn.microsoft.com/en-us/azure/ai-services/language-service/overview
3. Azure Synapse Analytics is a comprehensive data analysis and warehousing solution that allows you to process large datasets efficiently. The use case is ideal for querying and analyzing data across multiple sources, both on-premise and cloud-based - https://learn.microsoft.com/en-us/azure/synapse-analytics/guidance/proof-of-concept-playbook-overview
4. Azure Data Factory is a data integration tool for creating, scheduling, and managing data pipelines. it is useful for extracting and transforming data from ERP systems like SAP and JD Edwards for search indexing - https://learn.microsoft.com/en-us/azure/data-explorer/proof-of-concept-playbook
5. Azure API Management facilitates secure and efficient API management for data retrieval from various sources. The use case enables centralized access to data through a well-structured API layer - https://learn.microsoft.com/en-us/azure/devtest-labs/deliver-proof-concept

So, thinking about cost and licensing considerations: https://azure.microsoft.com/en-us/pricing/details/cognitive-services/

• Azure Cognitive Search: Costs depend on search units and storage usage. A free tier is available.
• Azure AI Language: Pricing is based on transaction volume and selected NLP features.
• Azure Synapse Analytics: Charges apply to data processed and compute resources used.
• Azure Data Factory: Pricing is based on data pipeline executions and volume.
• Azure API Management: Costs vary depending on API calls and the selected service tier.

For steps to get started:

1. Create a free Azure account with $200 in credits for 30 days and access to free-tier services for a year. You can sign up here - https://azure.microsoft.com/en-us/pricing/details/cognitive-services/
2. Utilize Azure DevTest Labs to build isolated environments for development and testing. https://learn.microsoft.com/en-us/azure/devtest-labs/deliver-proof-concept
3. Once the PoC is validated, analyze results and scale services accordingly.

This is a tested approach ensures an efficient, scalable, and cost-effective way to integrate natural language queries into your web application using Azure AI.

I hope this is helpful! Do not hesitate to let me know if you have any other questions.

---

Please don't forget to close up the thread here by upvoting and accept it as an answer if it is helpful.