It’s lurking behind all your favourite apps, serving up the type of content you prefer, filtering away your spam, optimising your travel routes and generally assisting you in living your life. We’re talking about AI, of course, and it’s been quietly evolving from basic machine learning to what we know today as generative AI (GenAI).
But it wasn’t until ChatGPT came along, with its ability to generate realistic and complex human-friendly [or readable] outputs, that AI was catapulted into the modern consciousness, significantly shifting AI’s perception and applications.
Microsoft has been central to AI’s fast-growing popularity. As one of the biggest investors in OpenAI (and therefore ChatGPT), they have been a driving force in bringing AI into the mainstream.
Naturally, Microsoft has AI scattered throughout its portfolio. For example, OneNote and PowerPoint leverage AI to enrich user interactions and design functionality. Embedded AI services and tools in Microsoft’s Azure AI platform offer developers and data scientists sophisticated tools like customisable machine learning models and form recognisers.
Microsoft CEO Satya Nadella at Microsoft Ignite 2023 said that Microsoft is “the Copilot company” and there is huge interest in all the various Copilot versions and AI-driven capabilities that Microsoft have developed.
At the top is Copilot for Microsoft 365, which is embedded in the Microsoft 365 apps you use every day to, in Microsoft’s words, “unleash creativity, unlock productivity and uplevel skills”. Copilot for Microsoft 365 can kick off the creative process in Word, help create engaging PowerPoint presentations, and support quick data analysis in Excel. In Outlook, it assists in managing email threads, while over in Teams, it can summarise meeting discussions and suggest items to action.
But for all its many uses, Copilot for Microsoft 365 comes with a hefty price tag, and a few conditions. While Copilot is generally available for small businesses with Microsoft 365 Business Premium and Business Standard licenses, these customers can purchase up to 299 seats. Commercial customers on E3 or E5 licensing agreements are not required to purchase a minimum number of seats. Copilot is also available to education faculty and staff. All Copilot licenses come at a cost of USD$30 per month for each user (as of December 2023), which is a sizable financial commitment, particularly for larger teams – or for small organisations where cost is more of a consideration.
But for all its many uses, Copilot for Microsoft 365 comes with a hefty price tag, and a few conditions. Microsoft Business and Enterprise customers can purchase Copilot for Microsoft 365 on a per-seat basis with no minimum number of users. Copilot licenses come at a cost of $44.90 per user per month with a 12 month commitment (as of February 2024). To enable AI for your workforce, this could add up to a sizable financial commitment. Do you provide Copilot for Microsoft 365 to everyone, or only to a select few?
If a Copilot for Microsoft 365 investment is giving you pause for thought, other AI-powered options are available. These could provide alternative approaches to examine specific use cases to enable and democratisation AI for your teams.
What is it? If you can get past the confusing naming conventions, Microsoft has rebranded Bing Chat and Bing Chat Enterprise simply as Copilot. (Not to be confused with Copilot for Microsoft 365) It’s a combined search and chat experience for Bing using ChatGPT models that comes with commercial data protection enforced for enterprise customers. There will still be a lot of people who refer to this as Copilot for Bing to avoid confusion.
Primary applications: Useful for generating content, analysing data, summarising documents, learning new skills, writing code, getting summarised answers to complex questions, and generally enhancing productivity and creativity in the workplace.
Security: User and business data remains within the organisation, Microsoft does not have direct access to the chat data, so nothing is saved. Data is not used to train the underlying AI models to maintain the integrity and confidentiality of your data.
Availability and cost: Available right now at no additional cost with Microsoft 365 E3, E5, Business Standard, Business Premium, and A3 or A5 for faculty licenses.
How to get started: If you are using any of the above licences, you can activate the service and start using this accessible Microsoft Copilot right away, including experimenting with prompt engineering.
What is it? Azure OpenAI is a suite of AI services and tools offering advanced AI capabilities like natural language processing and machine learning. Cognitive Services are a set of AI services and APIs that enable developers to easily add AI capabilities to their applications- think vision, speech, language, and decision-making. Think of these services as “off-the-shelf-AI” that you can use without having to be a machine learning expert or data scientist.
Primary applications: Use it to enhance customer interaction through intelligent bots, automate content creation, provide personalised recommendations, conduct sentiment analysis, and enable real-time speech translation. It is also widely used to develop AI-powered solutions for business analytics, customer service, and various other intelligent applications.
Security: User inputs and the resulting outputs are not shared with other customers or OpenAI, nor are they used to train OpenAI models or enhance other Microsoft or third-party services. No data persists within the system unless explicitly fine-tuned with user-provided training data, which remains exclusive to the user.
Availability and cost: Available through the Azure platform with the costs based on the specific service and usage levels. Options are available for pay-as-you-go and subscription models. In Australia, the services are generally available.
How to get started: To learn more, we suggest attending briefing sessions, exploring prototypes and watching demos that showcase real-world applications. That said, access is through the Azure portal, and the Azure OpenAI services are available in Australia.
Azure use case: AI-Assisted Document Analysis for Customer Support At Data#3 we are working with a customer whose customer service team has been sifting through dense 100 plus page documents to answer customer enquiries. To address this, we have fed the relevant documents to Azure OpenAI . The AI will then extract the relevant information for each inquiry and provide concise, accurate responses to customer questions. This will both accelerate response times and allow the customer support team to dedicate their efforts to more complex, high-value interactions. It’s just one example of how Azure OpenAI can be applied to specialised use cases to enhance service delivery and operational efficiency. |
What is it? A cloud-based platform designed for data scientists and developers to build, train, and deploy machine learning models at scale. There are a range of tools for all stages of the machine learning lifecycle, including data preparation, model training, deployment, and monitoring. Think of these services as a “build-your-own-model” approach where machine learning experts and data scientists can develop and publish their own AI models – then integrate those within applications and business processes.
Primary applications: Can be used for training and deploying machine learning models, orchestrating AI workflows, managing the entire lifecycle of large language models, and optimising model accuracy and performance.
Security: Includes governance, security, and compliance for running machine learning workloads, with built-in features for responsible AI. User data is not shared with others and is protected with options for double encryption.
Availability and cost: The service is available globally with consumption-based pricing. It can be either integrated with applications, used standalone or in conjunction with other Azure services. In Australia, the services are generally available.
How to get started: Access is via the Azure portal. First, you should consider your use cases and existing datasets to develop a pilot project. We also suggest exploring different services, setting up experiments, or even starting to build models.
Azure use case: Machine Learning to reduce operational overhead Data#3 helped an organisation who had undertaken an ERP transformation project. They regularly needed to categorise new product lines, however suppliers often had conflicting metadata associated with their stock items. To address this, our team developed and trained a machine learning model to categorise product items to the customer’s standard. For the migration of the organisations ERP, which had millions of SKUs and various labels for their product categories, hundreds of hours were saved and the solution continues to work behind the scenes to correctly categorise product items. It’s just one example of how Azure’s Machine Learning capabilities can be used to enhance productivity and reduce errors, allowing users to focus on higher-value activities. |
As Microsoft’s largest Australian business partner, with the highest level of competency across the Microsoft ecosystem and one of the few organisations on the Microsoft 365 Copilot Early Access Program, Data#3 is uniquely positioned to help you explore the options for embedding AI within your organisation.
With the speed of AI development and the breadth of Microsoft AI tools, we can assess your specific needs, existing infrastructure, and strategic goals through interactive workshops and data-driven insights, to recommend the right path forward for your business.
One thing is certain, the insights shared in this blog highlight a clear message: the time to act is now. Our team of Analytics and AI specialists are ready to help you decipher the data enigma for your organisation. Get in touch today to discover how we can help you develop a strategic plan that enables you to fully leverage analytics and AI in your daily operations.