Key Insights From the Team Lead Behind OpenVINO™ Notebooks: Part 2

Author: Paula Ramos — Intel AI Evangelist

Following Paula Ramos’ conversation with Andrei Kochin, Manager of the OpenVINO™ notebooks team, she had the opportunity to sit down with Ekaterina Aidova, AI Frameworks Engineer at Intel, who is on the forefront of creating new notebooks for developers. Ekaterina shared with Paula how she and her colleagues put their expertise together to build a baseline for new and experienced developers on OpenVINO.

Here are some insightful excerpts from their conversation:

Paula Ramos: Great to speak with you, Ekaterina. Please tell us more about yourself and your background.

Ekaterina Aidova: I joined Intel around five years ago as a software engineer and I currently work on the OpenVINO notebooks team covering model conversion.

My career started as a junior .NET developer. I was working on a project that was connected to development of an analytics platform for the analysis of social media sentiments. Similarly to Andrei, it was not part of my primary job description, but it related to artificial intelligence and data science. For example, the project involved neural networks for transcribing video to text and included OLAP cubes, a technology that’s known as the grandmother of big data. It was from there that I decided to continue my journey into the AI and data science field, which brought me to Intel.

Over the past five years, I’ve had many exciting experiences working across various teams on the OpenVINO project. I started out working closer to the runtime and devices and moved to different areas of the toolkit like plugins, tools, frameworks, and integration.

Today, I am part of the OpenVINO notebooks team, where I am continuing my journey to help AI developers through new and interactive user tutorials.

Paula Ramos: That’s some great experience you’ve had working throughout the complete OpenVINO ecosystem. Can you tell me more about the trends and technologies you are currently focused on within OpenVINO notebooks?

Ekaterina Aidova: Well, it’s no secret that generative AI has been experiencing a breakthrough for the past several years, generating more content like images, music, or text. There are a lot of examples of large language model (LLM) capabilities that can be used for chatbots, brainstorming ideas, writing code, or even producing essays. It’s a very interesting topic with the possibility to revolutionize our everyday routine.

The most interesting thing about all this is that it doesn’t necessarily require a lot of hardware resources to achieve, which we can demonstrate with OpenVINO, our notebooks, and a laptop.

Another trend is around multimodality, which has had a lot of success in the past couple of years. In my opinion, it’s one of the most powerful things in AI and comes with great social meaning. It’s my hope that my work will inspire people or yield practical results that can be used to improve the lives of different people. With multimodal models like BLIP or other models such as LLaVA, it is possible to analyze text, videos, or images, which can be used to help assist visually impaired individuals and provide them with the confidence and independence they need in their daily activities. It’s truly amazing what we can do with help of AI.

Paula Ramos: I agree that generative AI has the opportunity to change and improve lives. What are some of the popular generative AI notebooks available in the OpenVINO repository?

Ekaterina Aidova: First of all, we have different versions and support for the stable diffusion model, a popular model for generating images from text prompts or additional guidance through images as a starting point. We have been supporting it since its first generation and now we support the second generation of the model.

We also recently introduced the stable diffusion XL model, which can generate very realistic and interesting content based on user prompts.

Then we have models for text generation, and probably the most interesting use case is instruction following, which uses modern large language models like Dolly, Llamav2, or Zephyr.

You can achieve these tasks with OpenVINO, which you can find examples for in our OpenVINO notebooks. Additionally, we recently added a music generation model, and sound generation model, which generate music based on text prompts or generate realistic audio samples given any text input. It’s amazing that you can just open Google Colab, run our notebook, and generate audio content that you can share with your friends.

Paula Ramos: As a woman working in AI, what inspired you and how you can inspire other women in the field?

Ekaterina Aidova: It’s an interesting question. For me, AI is an area that requires you to solve various puzzles or challenges and allows you to collaborate with others in the community to solve those problems.

In web development, for comparison, where I previously worked, it was relatively easy to start and work because you could find answers to most questions on the internet. At times, it was intimidating that someone had already answered all the questions you had in mind, with some answers dating back several years. But web development is a stable field with a robust community.

On the other hand, the AI field is growing more rapidly. Every day you see new papers and new models, and there aren’t as many ready-to-use answers or established templates on how to think or approach problems. But there is something new to learn every day and a community you can constantly collaborate or share your experience with, which is very valuable. It’s not as difficult to enter this field as it may seem. All you need is a desire to learn. You can start by experimenting on projects right away or contribute to open-source projects. For example, OpenVINO has a special tag called “Good First Issue” for those who want to contribute but don’t know where to start.

Additionally, there are lots of online courses where you can start to learn the basics and continue your experience by reading blog posts and scientific papers. There are many resources available to help you become an AI engineer. It is amazing that this career track doesn’t require a specific gender or age. Anyone can try, achieve results, and contribute to this rapidly developing field — making an impact on everyday life and innovative technologies.

Paula Ramos: That is amazing, Ekaterina. I agree with you. This career doesn’t have a gender and doesn’t have an age. Anybody with or without experience can jump into this field. Thank you for sharing your experience and continuing to inspire developers working in this field.

Fun fact: To create this blog series, we used the whisper notebook. Try it in your next meeting!

And stay tuned for our next episode, with Samet Akcay, AI Research Engineer at Intel.

About Paula Ramos:

Paula is an AI Evangelist at Intel. She has been an AI enthusiast and has worked in the computer vision field since the early 2000s. Follow her on Medium for more insights.

Notices & Disclaimers

Intel technologies may require enabled hardware, software, or service activation.

No product or component can be absolutely secure.

Your costs and results may vary.

© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.