Lesego Makaleng
The Microsoft Build AI Day on 18 March 2024 brought together global and local industry leaders with cutting-edge perspectives to explore how AI can help organizations drive innovation, efficiency, and growth. Different capabilities from Integrove participated in this event to learn more about leveraging AI in our day-to-day projects. In this post, we will go through the events of the day and discuss how Integrove leverages AI tools.
Microsoft Build AI Day
The Microsoft Build AI Day started with the Microsoft Partners Showcase. Different partners came to showcase their organisations, software and hardware. Many of them engaged with us on how they integrated AI into their operational processes and software offerings. This shows that the wave of AI has swept over the entire industry. It is critical that organisations in the technology space adapt in such a way that they can integrate AI into their operational processes to be more efficient.
Introducing Microsoft’s GitHub Copilot
After the showcase, we gathered in the main hall for the welcome session. In this session, the managing director of Microsoft in South Africa gave us an overview of Microsoft’s offerings around the AI space and spoke about how AI is meant to simplify our daily tasks instead of replacing us.
Immediately after this, we were introduced to Azure Machine Learning where Laurent Bugnion took us through creating our own Large Language Model (LLM) and talking about how we can use this tool in our daily work life. The main idea behind LLMs is that it uses deep learning techniques and massively large data sets to understand, summarise, generate, and predict new content. We also delved into generative AI specifically architected to help generate text-based content. What was interesting about the session was that we found that LLMs can help us enable a conversation with a user in a way that is typically more natural than older generations of AI technologies.
Towards the end of the session, they officially introduced GitHub copilot. GitHub Copilot is a code completion tool developed by GitHub and OpenAI. Surprisingly, only half of the room had interacted with GitHub Copilot even though this offering has been out for over two years. They took us through a practical example of using GitHub Copilot by building a simple Python application. There are multiple ways you can use GitHub copilot to build applications. The most common way is to prompt it by describing what kind of code it should write. Then you can either accept or decline the suggestion it makes.
There are other features that GitHub Copilot can be used for, such as:
- fixing code errors,
- optimizing code,
- writing tests,
- adding documentation,
- and explaining code or even an entire project.
GitHub Copilot chat allows you to have a conversation with Copilot and ask it questions that are code-related, explain pieces of code, or explain the entire project to you. This is a useful tool, especially if you’re starting out on a project.
Copilot at Integrove
In the Enterprise Application Development (EAD) Capability at Integrove, we have been using GitHub Copilot for over a year. We use it to deliver quality solutions in a shorter period. GitHub Copilot has allowed us to focus on the business logic of our enterprise applications while using it to write utility functions, unit and integration tests, and add code documentation. This allows us to deliver well-documented and tested solutions to our clients.
Hands-On Workshop
For the final session of the Microsoft Build AI Day, we headed over to a hands-on workshop where we would put into practice everything we’d learned throughout the day. We had to build a rock, paper, and scissors game in Python by making use of GitHub Copilot. It became a competition in the end to see who would build the best game. The session was focused on using GitHub Copilot rather than on your knowledge to see how much you could extract from Copilot.
In the end, everyone showcased the different solutions they came up with. From the Integrove delegation, in 20 minutes, we were able to develop the game and add some voice recognition features to make it more interactive. A lot of the work was put on GitHub Copilot to figure out how to do voice recognition in Python, and after fixing some missing libraries, we were able to showcase this to the rest of the people in the session.
Looking back on the Microsoft Build AI Day
Overall, the Microsoft Build AI Day was informative, and we made many wonderful connections with people in our field. We discovered how crucial it is to include AI in our normal company operations to increase productivity and provide our clients with high-calibre solutions. At Integrove, we’ve used AI productivity tools like GitHub Copilot to aid in the development of enterprise-level solutions.