Hackathon participant

Hot to hack

24 February 2020Mateusz Chrominski

How many hot ideas can you develop in just 8 hours? As our latest Hackathon proved, quite a lot!

There's nothing quite like the buzz of a Hackathon. A room full of passionate developers, all hell bent on winning the coveted prize for the best idea. An expectant but critical jury, ready to pick the holes in an idea, if they exist. And a focus that produces a level of innovation that you just can’t tap into in the regular routine. We run at least 2 major Hackathons a year, changing the theme every time to give developers a different perspective and to challenge them to dig deeper into their shared expertise.  

For me, our hackathons are great opportunities to work with various experienced people, experts in given areas. You learn a lot, discuss ideas and think outside the box.

Marcin 

Our latest Hackathon had one key goal: to bring as much innovation to our core areas of focus for 2020; namely Content as a Service and Experience as a Service. 

83 Cognifiders, divided into 17 teams spent an entire day planning, exploring and coding new solutions for our AEM projects and products.

A deep dive into assets

Customer attention in the experience age is a valuable commodity. According to a BBC study, the average attention span dropped from 12 seconds in 2000 to 8 seconds in 2017. Rich and engaging imagery undoubtedly attracts that attention and helps to keep consumers focused. But constantly providing eye-catching experiences is not an easy task and puts businesses under pressure to maintain huge asset libraries, making them intuitive to use and easily searchable.  

However, with data storage exceeding terabytes of content and imperfect metadata description processes, searchability becomes a true challenge. This was an area that a few of our teams took on for the day.

A couple of teams came up with innovative enhancements to the search function in AEM Assets by extending the concept of asset tags. One involved an integration with an ontology (the interpretation of a group of ideas within a specific domain that defines the interrelationship between those ideas), defined in a large data set by Princeton University. This extended a search beyond a specific tag to other related keywords such as synonyms and tag hierarchies. For instance, in a keyword search for ‘car’, the integration would automatically include a related word like ‘vehicle’ and vice versa. 

Another team created a new capability that helped to build tag hierarchies directly in AEM. Other teams tackled the issue of duplication of assets and another created an integration between AEM, Content Fragments and voice recognition. 

Hackathon participants

AEM as a Cloud Service 

Since AEM as a Cloud Service is Adobe’s new kid in on the block, several teams looked at what could be created around this offering, that gives organisations hyper-fast, SaaS-like agility. With more and more out of the box capabilities available in AEM, less effort is needed to actually build an experience. But the more experiences that organisations create on a platform, the tougher the actual management of it is. This new accelerated approach to creating, managing and delivering content gives us the opportunity to reassess the tooling we use and processes we follow. 

Some of the solutions focused on shortening the development process and enhancing the Continuous Delivery / Continuous Integration setups. One team, looking at atomic design, explored the difficulties around content governance in vanilla AEM. The proposed solution involved a clever integration with Storybook – a React library. 

One team prototyped an automated tool that can quickly estimate the cost of building a website based on designs. The technology scan’s the design, identifies common building blocks and then crawls through our catalogue–knowledge base that links building blocks with a development effort estimate. Another great application of AI to save time and effort! 

AI-based solutions are the future 

Fantastic technical craftsmanship was demonstrated by the team that developed a Machine Learning Python application. Based on a challenge that one of our clients is currently facing – the recognition and categorisation of various personas in their large dataset of assets.  The solution worked surprisingly well with around 85% accuracy. That was voted the Hackathon’s hottest by our jury! 

The jury appreciated the use of a variety of new technologies, clarity in the problem definition and the presentation of a working solution by the end of the day.

Albert Cenkier, Technology Director

Just the beginning

Each year the level of complexity in experience technology increases. The Hackathon provides our teams with the opportunity to marry the demands of our work with the fast-changing technologies that are on top right now and bring their own innovative thinking to problem solution.

Hackathon team

With another Hackathon in the calendar for April, our developers are already thinking about what they can bring to the table this time. Stay in touch to learn more and see how we bring our innovations to market in the future. 

Author: Mateusz Chrominski
Published: 24 February 2020
Tags:
AIAdobehackathonculturetechnology
 

People in or team love to share their experience. Explore our blog

Job Opportunities

We're always looking for new faces to join the Cognifide family. Take a look at our available jobs.