3 Innovative Ways Developers Are Building with AI

AI is reshaping industries and unlocking new opportunities for startups and enterprises to grow. Azure AI is at the forefront of this transformation, offering a comprehensive suite of tools that empowers you to build, manage, and deploy AI applications at scale. Programs like Microsoft for Startups Founders Hub which include access to Azure AI models and tools enable founders to turn innovative ideas into reality.

This past April, Microsoft hosted a multimodal themed Generative AI Hackathon on Devpost, where developers showcased their projects built with Azure AI. Participants of the hackathon used Azure AI to create innovative solutions across education, accessibility, fashion, media, and more. There were a wide range of use cases addressing real-world challenges responsibly:

1. AI as a Catalyst for Enhanced Learning

Project: ChatEDU | Industry: Education

ChatEDU exemplifies how AI can transform education by creating personalized, interactive learning experiences. The project moves beyond simple automation, offering a dynamic copilot that evolves with students, fostering curiosity and skill-building.

“Choosing Azure AI was a game-changer for us. Its robust suite of tools enabled us to transform diverse educational data materials into an intuitive and interconnected product. This has been pivotal in creating ChatEDU, a personalized, multimodal educational tool that enhances student learning experiences” – Jason Hedman | Founder | ChatEDU

The solution uses Azure AI Services like Azure Document Intelligence to understand and process knowledge files uploaded by students. This information from multiple data sources (plain text, images, videos, PDFs) is stored in Azure Cosmos DB. These files are then analyzed to extract topics and connections to create an interactive learning pathway that students can utilize to work towards learning objectives.

2. AI for Accessibility and Inclusion

Projects: GARVIS, Gino.AI | Industries: Technology, Media

GARVIS redefines accessibility by empowering users to navigate visual tasks without relying on verbal descriptions. The app utilizes models from Azure OpenAI Service to detect and analyze objects and environments, replacing text-based instructions with intuitive visual demonstrations. Using the Unity Engine paired with Microsoft Copilot to enhance coding productivity, this application makes advanced mixed reality assistants available to all. A complex space simplified with the ease of using multiple Microsoft products together.

Another example of seamless integration is Gino.AI, which makes information accessible and consumable in a way that caters to your learning style. Azure AI Search is used to index and efficiently query information from the ‘Mind Base’ knowledge space, and with Azure AI Vision, the system can extract text from images, enriching the content available for users. To ensure responsible AI use, Azure AI Content Safety filters out any harmful material from chatbot responses, summaries, podcast scripts, and other sources.

3. Leveraging AI to Maximize Existing Data

Projects: FarmFundAI, Winewiz, Docsy, Therapute | Industries: Agriculture, Retail, Healthcare

FarmFundAI demonstrates how AI can unlock new value from existing information available, helping farmers secure funding by showcasing sustainability and crop yield potential. By analyzing satellite and drone imagery via an app built with Azure AI Studio and Azure Functions, users can uncover insights like crop health that are crucial for investment decision-making.

This concept of utilizing existing publicly available data is also shown through Winewiz which simplifies the process of selecting and gifting wines by offering AI-powered conversational consultations and gift card creation. Leveraging Azure AI Speech, it provides interactive communication through the Speech-to-Text service, enhancing the user experience and making the selection process intuitive. Similarly, Docsy saves time for healthcare professionals and enhances the accuracy of patient records via audio data that is processed through Azure AI Speech for transcription. Within the same healthcare realm, Therapute is a physiotherapy platform that uses Azure Machine Learning to offer personalized rehabilitation guidance 24/7.

Building Multimodal and Cross-Domain AI Applications

As shown in the use cases above, Azure AI is versatile, supporting multimodality across industries and domains that require diverse data types, such as text, speech, images, and video. Leveraging the Azure AI model catalog is one way to make the most of this flexibility, where you can select the right AI models for specific needs and constraints such as price, performance, and customization. Multimodal models like GPT-4o via Azure OpenAI Service and Microsoft’s very own Phi-3-vision are also part of the model catalog that you can try out today.

It’s exciting to see that developers are utilizing Azure AI in conjunction with other Microsoft products. Talk to Listen is an example of this, where developers used Azure AI with GitHub Copilot to create a high-quality app bringing characters to life, all while saving time. We are thrilled to see what developers will continue to build and innovate.

Feeling inspired to start your own project? Start your journey by exploring helpful tutorials and resources:

Learn

Get started with our Generative AI for Beginners course

Get started with Phi-3 using our cookbook

Learn about generative AI use cases on Azure

Learn how to use Azure AI Studio on Microsoft Learn

Connect

Join a passionate community of builders on our Azure AI Discord

Learn, Connect, and Build with us on Microsoft Reactor

Have a startup? Sign up for Microsoft for Startups for exclusive access to mentors, Azure credits, and more.

Build

Build with Azure AI Studio