From Aircraft Engineer to AI Data Specialist – an interview with Gosia Los-Brzozowska

From aircraft engineer to AI data scientist, Gosia Łoś-Brzozowska’s career is anything but typical. At Spyrosoft’s Białystok office, she’s turned a knack for data into a force for innovation, working on everything from generative AI to predictive maintenance. In this candid chat with Karolina Spryszynska, Head of Employer Branding, Gosia shares her path, the challenges she’s tackled, and what it means to be among the Top 100 Women in Data Science.

Gosia, what did your path to AI Data Scientist look like?

Well, it all started in high school; I wondered what kind of studies would be interesting for me. Once, I went to a mechanics lecture and understood nothing; all the Greek letters, formulas…But despite that, I felt like something was leading me that way, so I decided to follow that path. After university, I started working as an aircraft engineer in the design and manufacturing areas. My tasks included redesigning, evaluating field experiences, and approving manufacturing processes. Every decision had to be made based on data. This is where my adventure began as a Data Specialist.

Could you share your opinion about the most interesting data-driven AI solutions project that you took part in at Spyrosoft?

Choosing the most interesting one is no easy task! At Spyrosoft AI, we work on a variety of projects, from cost optimisation to autonomous vehicles. Recently, the hottest topic has definitely been gen AI. Our projects aim to develop data driven AI solutions that leverage advanced computational methods to enhance decision-making and improve operational efficiencies.

My first professional experience with generative AI was a thrilling dive into the future of software testing. At Spyrosoft, I was developing an intelligent assistant for generating test scripts for QML applications using machine learning.

Imagine this: A tool that not only understands the intricacies of QML but also autonomously crafts precise, efficient test scripts. Our goal was to develop an AI that could read through the QML code, comprehend its structure and functionality, and then generate comprehensive test scripts, all with minimal human intervention. It was a thrill – working with cutting-edge AI models, brainstorming and conducting experiments. It’s the perfect playground.

Have you faced any challenges in your recent projects? Tell us what they were about.

Challenges are the daily bread in AI. I haven’t worked on the project without them. Automating and optimizing business processes is crucial to overcoming these challenges. Most often, we face meeting financial constraints, which requires close budget monitoring, automation per-custom needs, often of specific if not unique use, and finally, managing client expectations. That’s the GIGO effect – Garbage In, Garbage Out. No magic. You need data, computing power and highly qualified people to incorporate AI.

Anyway… There is a resolution to each challenge. And the outcomes are so satisfying when supported by a robust data management infrastructure.

What are some common pitfalls and risks in planning a data science project with data management infrastructure?

There is a joke:
– What is important when buying an apartment?
– Only three things: location, location, location.

If it goes about planning a data science project or any project, I could rephrase the joke and say: Communication, communication, communication. Adapting communication techniques is crucial for resolving the problem and ensuring all involved parties understand the importance of data quality. Moreover, even with technologies such as ChatGPT, some time is needed to research, develop, test, and deploy.

Can you share an example of a successful data science project you led using machine learning? How did you leverage data to provide valuable insights?

I led a predictive maintenance project for a newly introduced aircraft fleet, ultimately, a few thousand units. The goal was to rank the engines in terms of overhaul need, thereby preventing unexpected failures and minimising downtime through comprehensive data analysis. This project used a Cumulative Damage Model (CDM) learning from design, production, operational and environmental data. The tool provided not only valuable insights into engines’ health, but also contributed to more efficient and cost-effective maintenance practices. Additionally, the project helped enhance human decision making by integrating advanced technologies to optimize maintenance workflows.

You have also been named one of the Top 100 Women in Data Science this year; congratulations! What does this mean to you?

Thank you for acknowledging that! Being recognised as one of the Top 100 Women in Data Science this year is a significant honour. It means a lot because it reflects my hard work and dedication, and it proves it’s not only my feeling. It is incredible to be included among talented and fantastic women, which also demonstrates the inclusive culture in IT.