Azure's generative AI is a powerful and versatile technology that can help users to create and deploy intelligent applications that can generate content, insights, and solutions from your own data. It can be applied to almost all industries and domains, such as education, healthcare, media, entertainment, gaming, marketing, public sector and more. Azure's generative AI can help users to automate repetitive tasks, enhance creativity, and solve complex problems. GenAI can be used as a Copilot or a Custom Copilot (bespoke build), depending on the level of control and customization that the user needs.
- Copilot: This is the default mode of GenAI, where a user can enter a prompt or a partial text and GenAI will complete it with relevant and coherent content. It uses a general-purpose model that can handle a wide range of topics and domains. The Copilot mode is useful for tasks such as writing emails, blog posts, social media posts and product descriptions.
- Custom Copilot: This is an advanced mode of GenAI, where the user can create their own models by fine-tuning general-purpose models on their own domain specific data. The user can also specify the style, tone, format, length, and other parameters of the generated content. The custom Copilot mode allows the user to train a specialized model that can capture the nuances and specificities of their use case. The custom Copilot mode is useful for tasks such as creating personalized and targeted content for specific audiences or scenarios.
What are some potential use cases of Azure's generative AI?
We have been working with many customers and industries sectors and came across numerous use cases. Here are some examples of how Azure's generative AI can help users in different scenarios.
Manufacturing: As the industry manufactures finished products or parts rather than services multiple use cases can be seen:
- Many organisations have historical and technical documentation. An organisation may want a way to surface useful information from the documents and query it in natural language. This not only reduces management administrative efforts, reduces labour costs and overhead costs (for physical storage to store files). An internal Copilot can be created for employees to ask questions in natural language and get back an answer. This involves deploying a solution to extract relevant contextual information from a Knowledge Base. Using this tool a custom Copilot can be made to answer your organisations specific questions.
- A Copilot could provide valuable assistance to manufacturers by suggesting designs and recommending optimal materials. These recommendations consider cost, sustainability, and durability considerations. For example, Rockwell Automation, a leading US provider of industrial automation technology, leverages Microsoft Copilot within its FactoryTalk Design Studio. Copilot assists engineers by generating code through natural language prompts, automating routine tasks, and enhancing design efficiency (click here for further reading on this use case).
- Generative AI can also be used to build cloud-native systems to improve efficiency by gaining real-time insights on production lines or industrial equipment. This moves from batch processing to real time allows for an improved customer experience.
- Copilot can enhance innovation and operational efficiency in any organisation. For example, Siemens is integrating its Teamcenter software (used for product lifecycle management) with Microsoft Teams and Copilot. This solution allows for:
- Production operatives to use their devices to report design concerns in natural language, the summary of the reports sent in by the production operatives. GPT 4v can assist in terms of analysing the images and visual data. This helps to detect defects or inconsistences in the production line(click here for further reading on this use case)
Retail: It’s essential for retailers to standout by bringing appropriate products to customers on hand at speed. This can be accelerated with the help of generative AI. Some use cases seen in this industry include:
- Personalised product recommendations - to maximise sales, tailored advertising and marketing is used to recommend products based upon a customer’s purchase history, preferences, and behaviour to aid the alignment of promotions to the customer. Azure's generative AI can help marketers and advertisers to create and test content on various user groups. Custom Copilots can be used to enable users to chat with the system database to find appropriate products that might be best suited for their needs. This chatbot allows a user to search the retailer’s database in natural language to obtain a result.
- Forecasting & inventory management. Generative AI can help retailers predict future demand and optimise inventory levels, reducing cost and waste. This can be done by analysing historical data, market trends, and external factors to predict future demand more precisely. This accuracy helps retailers optimize inventory levels, preventing stockouts (which disappoint customers) and overstocks (which lead to waste).
A customer example of how generative AI is used in retail, is Estee Lauder click here for further reading on this use case.
Public Sector: Generative AI has the potential to revolutionize how challenges are addressed in the public sector. To increase the efficiency (more than 30%) of Government Departments aligning to the Cabinet Directives can utilise generative AI.
- Using Azure Open AI, chat Bots can provide better customer service to provide citizens with information e.g. Gov.UK. Chatbots would be able to understand and interpret natural language queries from citizens. NLU (natural language understanding) models would process user input, extract intent, identify relevant entities and relay the answer back to the user. Chatbots can provide and share knowledge internally with more people and in some instances might be able to explain information with trends in data that might not be able to be detected at first glance.
- Automations, Applications and Processes infused with the Azure OpenAI service can unlock high levels of efficiencies in key areas such as call centres, citizen services and borders. Azure Open AI can simplify call centres e.g. HMRC Tax Helpline, automate manual processes e.g. DVLA Driving Licence Applications. Phone calls between agent and customers may be recorded and stored and later Azure AI speech can transcribe the audio files asynchronously while identifying different speakers, languages and sentiment (click here to read a customer use case).
Financial Sector: (Finance industry) encompasses institutions or services involved in the management of money. Generative AI has potential to transform the industry.
- In portfolio management, Copilots can assist portfolio managers by analysing market trends, suggesting investment strategies, and providing real-time insights. For instance, a custom Copilot could monitor stock prices, analyse financial news, and recommend adjustments to investment portfolios.
- Augmenting Human Capabilities Through Automation, this allows for more focus on strategic activities. Extraction of insights from documents and summarisation can be done utilising Azure generative AI capabilities. It has the capacity to analyse and synthesize vast amounts of financial documents, such as reports, contracts, and regulatory filings. With the ability to extract information and identify patterns to aid in the prevention of fraud. It increases the efficiency of the organisation. In claims, it would help claim handlers and claim adjusters better manage customer interactions and help reduce fraud.
The potential use cases of Azure’s generative AI are vast and continually evolving, demonstrating its versatility and power in addressing industry-specific challenges and enhancing operational efficiency.
In the next article we will discuss about how to begin building custom Copilots.https://techcommunity.microsoft.com/t5/ai-ai-platform-blog/the-evolution-of-genai-application-deploy...
@Paolo Colecchia@arung@Stephan Rhodes @Renata Bafaloukou @Morgan Gladwell
