Open Source LLMs Dev Day puts early career AI experts in the spotlight

IBM TechXchange Dev Day: Open Source LLMs on April 22, 2025, gives you hands-on experiences that are specifically designed for AI developers, data scientists, integration developers, administrators, and engineers. Instructor-led labs showcase generative AI-based use cases to benefit your gen AI ecosystem. You'll learn how to build AI agents and assistants during hands-on workshops that delve into generative AI and automation technologies using IBM Granite open source AI models.

Here, we introduce two of our Dev Day presenters and preview the kind of valuable hands-on experience you'll get when you register and complete their sessions. And if you miss Dev Day on April 22? You can still sign in and watch the session recordings. Don't miss this opportunity to build your AI skills!

Hailey Quach, Data Scientist

How did your journey at IBM begin?

It started with an internship in Sales Enablement under the IBM Skills Network team. The interview process was a mix of technical and communication challenges—I was asked to walk through past projects and explain core machine learning and data science concepts in simple, accessible language.

What stood out was IBM’s emphasis on curiosity, clarity of thought, and communication, not just pure coding ability. They weren’t just evaluating what I knew, but how well I could teach what I knew.

What team are you on, and what kind of work do you do?

I’m part of the IBM Skills Network team focused on learning content development for AI, data, and automation. Our mission is to make AI more accessible to IBMers and external partners through education, enablement resources, and live events.

On a typical day, I’m prototyping AI use cases, writing tutorials, building demo apps, or putting together workshops. I work with tools like LangChain, LangGraph, CrewAI, and watsonx.ai—all in service of making advanced AI topics digestible and hands-on. It’s a sweet spot between education, product experimentation, and storytelling.

You recently spoke at an IBM TechXchange Dev Day. Tell us about that.

In January, I hosted a hands-on workshop for IBM TechXchange Virtual Dev Day called Agentic AI: Build a multi-agent app with CrewAI, Llama 3.2, and Gradio. It was a live coding session where I showed how to use the Llama 3.2 90B vision-instruct model and a multi-agent system built with CrewAI to create an AI Nutrition Coach. The app could recommend recipes and estimate calories/nutrition from food images.

Surprisingly, it became the most-watched session of the entire event on the IBM Developer YouTube channel.

What advice would you give to someone who wants to thrive in this kind of role?

1. Be resourceful: You won’t always have a playbook. Learn to figure things out and make progress in ambiguity.
2. Stay curious: Don’t obsess over perfection. It’s okay not to know everything. Keep asking questions, challenge ideas, and treat every project like a chance to learn.
3. Teach what you build: Whether you're explaining an agentic workflow to a junior dev or a senior architect, being able to communicate clearly is a superpower.
4. Engage with the community: IBM has a rich, open culture. Connect with other builders, share your work, and learn from theirs.

What’s next for you at IBM?

In April, I’m presenting at the IBM TechXchange Open Source LLMs Dev Day. It's a deep dive into Agentic RAG architectures, showing how developers can build intelligent, retrieval-augmented systems using LangGraph, Docling, watsonx.ai, and open-source LLMs.

As I write this, I’m also preparing to join a new internal team focused on AI enablement, where I’ll design modular training tracks on applied AI and automation topics like multi-agent orchestration, memory integration, and evaluation strategies.

Any final thoughts for aspiring data scientists or AI builders?

IBM is an amazing place to learn by doing, especially if you love building and teaching. You don’t need to come in as an expert, but you do need to be proactive, experimental, and community-driven.

Anna Gutowska, AI Engineer

What's your role at IBM?

I'm an AI Engineer and Developer Advocate for ibm.com. My work is dedicated to empowering developers to effectively leverage advanced AI technologies in a user-friendly way regardless of experience level. The technical content I produce consists of tutorials, explainers, videos, and conference presentations with a focus on building agentic systems. Apart from creating technical content, I am also involved in developing internal tools to streamline our team’s manual processes as well as improve the quality of our content.

How long have you been with IBM?

I joined IBM in June of 2024 and it has been an incredible learning experience and growth opportunity. In this time, I have not only furthered my technical writing skills but also my software development skills. This role requires me to stay informed on the latest AI advancements and often involves working with experimental technologies. The problem-solving and exploratory aspects of my work are what excite me in the day-to-day. But the feedback I receive from developers working directly with my content is what brings me the most joy.

What's your top recommendation for current and aspiring developers?

For developers looking to build their skills, I would recommend staying informed and working hands-on with the latest AI advancements, particularly in agentic AI. There are endless opportunities to leverage AI agents in impactful ways and knowing how to build custom agents is key.

What are you presenting at IBM TechXchange Dev Day: Open Source LLMs?

I'm presenting alongside Ash Minhas, the Technical Content Manager for ibm.com. In our talk, we will demonstrate our generative AI-powered personal stylist. The stylist leverages the IBM Granite Vision 3.2 LLM for processing image input and the Granite 3.2 Instruct model with the latest enhanced reasoning capabilities for formulating customizable outfit ideas. Our presentation will explore the logistics of building the application and will feature a live demo.

What's the most important takeaway from your talk?

This application is unique in that it showcases how to use the latest two Granite models in a complementary way for handling both image and text inputs. Attendees will find that there is customization available to produce outputs personalized to the user’s occasion and objective. This use case applies to many of us in our daily lives and there are lots of opportunities to build off of the application. I highly encourage you to listen to our talk and check out the latest that Granite has to offer!

Register for the live sessions or check out the replays

IBM TechXchange Dev Day: Open Source LLMs – Asia Pacific happens April 22, 2025. Register now to join Hailey, Anna, and other expert presenters for their hands-on sessions. And if you miss the live sessions, you can sign up to check out the session recordings. Don't miss this opportunity to take your LLM and AI skills to the next level!

Register for IBM TechXchange Dev Day: Open Source LLMs – Asia Pacific Edition