Twitter LinkedIN Xing facebook

HamburgAI#9 Recap: Data Minimalism, Machine Learning Pipelines & AI in Media Content Creation

For the third time this year collectAI hosted the AI community event HamburgAI on the premises of WeWork Stadthaus.

Around 80 guests from startups, companies and academics joined us on July 31st to listen to presentations from hybrid cloud data service company NetApp, Germany’s largest e-commerce retailer OTTO and video optimization start-up Aiconix. Together we discussed topics related to Artificial Intelligence and Data Science. Some highlights of the talks are summarized in this post.

HamburgAI will be back for its anniversary event in autumn this year. Details will be announced in due time. Subscribe to our newsletter to stay tuned.

Machine Learning Development Insights: Optimizing the AI-Pipeline with Kubeflow

Solutions Engineer Steve Guhr and AI Solutions Architect Muneer Dedmari from NetApp, a California based tech company listed in the renowned Fortune 500 ranking for the past seven years, opened the event with their talk.

The two discussed their experiences on building platform services for managing AI and ML-pipelines in an efficient manner: “It’s all about automation and self-service, whether it’s in the cloud or built as an on-premises infrastructure”.

For many companies, managing data seems to be a very crucial task, yet often underestimated challenge. Therefore, their talk highly focused on the process of automating data pipeline creation using UI-based Kube-Flow solutions.

Download their presentation slides here.

Enhancing Machine Learning Results through Data Minimalism

Otto Group has hit the ground running, entering the digital age faster than many competitors. With online sales of 7.7 billion euros, they have become Germany’s largest e-commerce retailer. Michaela Regneri, Ph.D., Senior Expert AI & Cognitive Computing, credited Otto’s successful digital transformation for their phenomenal assimilation to the new age.

In her talk, she tackled the topic of data minimalism. In a nutshell: “Everything is better with bacon” might be a sentence many of us agree, however, the data should not be treated as such. Michaela explained why minimalistic data should always be the preferred choice for any AI and underpinned her arguments with data from OTTO.

Download Michaela’s presentation slides here.

Creating Videos with Artificial Intelligence

Aiconix, a start-up capitalizing on Machine Learning in the field of video optimization, was represented by its CEO Eugen Gross. In a panel discussion with collectAI’s Senior Machine Learning Engineer Omar he pointed out the vast potential of using Artificial Intelligence in media content creation.

With decades of experience in videography, his expertise lies especially in the creative process. Talking about the challenges of entering the market, he explained that little has been done so far in terms of applying machine learning in the process of video creation. Going further, he shared insights on how aiconix applies semantic analysis, facial recognition, celebrity recognition and many more to achieve better results in video production.