iQmetrix Expertise in the Wild: An Immersive Workshop on Semantic Kernels and AI in Action

As a technology professional who leads the tech side of iQmetrix’s Data and Analytics team, I’m always looking to use expertise and knowledge to help the wider tech community. 

Being an early user of an open-source Software Development Kit (SDK) called Semantic Kernel, which I have also contributed code to, it’s important to me to help others learn to use it and apply it to the real world. 

There is a lot of official documentation available, but users can get easily lost when they dive in. I wanted to find a way to show how it can help extend AI capability in realistic use cases. So, on February 27, ITAs — a programming organization of which I am one of the founders — hosted an immersive workshop titled Pie & AI in Action: Semantic Kernel Workshop.” 

As we gathered at the Microsoft Canada HQ office in Toronto, the winter chill was no match for the warmth of the conversations that filled the meeting room. It was a workshop, yes, but it felt more like a gathering of old friends, all passionate about the potential of semantic kernels and bringing AI into production. 

What did the workshop focus on?

Semantic Kernel is an SDK that lets you easily combine AI services such as OpenAI, Azure OpenAI, and Hugging Face with conventional programming languages like C#, Java, and Python. 

The focus of our session was to learn and discuss how semantic kernels can be applied to solve real-world problems, with a particular emphasis on two common use cases: AI Agent and Retriever-Reader (RAG) models. 

AI Agent:

AI agents are intelligent systems designed to perform specific tasks or interact with users autonomously. Examples include chatbots, virtual assistants, or recommendation engines. 

Use cases:

  • Customer Support Chatbots: AI agents handle customer inquiries, troubleshoot issues, and provide relevant information.
  • Personal Assistants: These agents schedule appointments, set reminders, and answer questions.
  • Recommendation Systems: AI agents suggest products, movies, or music based on user preferences.

Retriever-Reader (RAG) Models: 

RAG models combine generative language models (like ChatGPT) with an information retrieval system. This enhances responses by grounding them in relevant data. 

Use cases:

  • Search Q&A: RAG models retrieve precise answers from a large body of documents.
  • Chatbots with Content Retrieval: They pull relevant information to provide accurate responses.
  • Advanced Chat Applications: RAG enables context-aware conversations. 

Who attended the workshop?

The attendees were as varied as the applications of AI itself. There were technology startup founders with visions of changing the world, programming researchers who spend their days pushing the boundaries of what’s possible, and even a finance professional whose daily language is numbers and probabilities. Each brought their own set of unique challenges, goals, and pieces of the puzzle, and together we started to see a bigger picture. 

Despite the diverse programming backgrounds of attendees, the well-designed semantic kernel labs ensured that all attendees — whether they were familiar with programming languages C#, Python, or Java, or something else entirely — could actively participate and follow along. The format fostered inclusivity, encouraged questions and discussions, and allowed participants to share their thoughts and learn from each other’s experiences. 

The discussions that unfolded were as rich and varied as the participants themselves. Questions led to deep-dives into topics that went beyond the surface-level understanding of AI. It was a reminder that sometimes, the most valuable insights come from simply talking things through with others. 

How did the workshop go?

Throughout that evening, what stood out to me was the participants’ eagerness to explore how semantic kernels could integrate into their projects. It transcended mere comprehension of the technology; it was about envisioning its practical application in the real world. The labs were not simply exercises; they represented the initial strides towards genuine, tangible innovation. This applied whether one was developing a copilot for personal finance assistance or a business intelligence agent. 

It was also great to see that we had technological smooth sailing. One of the biggest hurdles in any technical workshop is the it works on my machine” syndrome. To sidestep this, I provided a hosted lab environment with individual logins. It was a simple yet effective solution that allowed us to focus on the learning experience rather than getting bogged down by technical snags. 

Feedback after the event, on both the practical approach of the workshop and its content, was unwaveringly positive. Fras Wasim, Associate Research Scientist at SCIEX, wrote in response to the session, The event was incredibly insightful and sparked some new ideas for me. I’m excited for the next one!”

The event was a microcosm of the tech community at its best: diverse, collaborative, and forward-thinking. It’s evenings like these that remind us why we got into this field in the first place — to learn, to share, and to create something meaningful together. 

Thank you to everyone who came out and made the workshop such a memorable experience. Your passion and curiosity are what drive these events forward, and I can’t wait to see where our discussions lead us next. 

Want to see iQmetrix’s experts sharing innovation and inspiring technology solutions? Sign up for updates on this year’s SummitX event — registration opens soon!