The good and the bad of Burda Media’s hackday
Last weekend I participated in the Burda Media hackathon. The Hackathon/Hackday topic was “the future of finance”. I have to admit that I had always wanted to go to a Hackathon and try out my programming-cool-things-under-pressure skills. But since my arrival in Germany I did not have the chance or it was not my priority.
I am not familiar with Fintech at all, so I thought it was also a good opportunity to learn a bit more about the topic. I contacted the (in)famous Karim Jedda and during the social event we teamed with Moritz “Kandinsky” to form our designerless 3-man army. We came up with a a small hack called Robo.Tax. The ideas was to to try to help normal people with the Tax declaration by analyzing the documents and receipts you receive via e-mail (or the ones you take a picture with your phone.). We used the Gini API to do that. It sort of worked although the Gini API does not yet provide the whole metadata needed to extract the taxes, we did some hacks to demostrate our idea.
You can see our hacky, french-accented, bad recorded demo video here:
I really liked the location, the HvB Forum, a massive building with an old facade but with a really modern interiors. There where some issues regarding getting energy in some of the location of the building and the HDMI cable was not easily accessible at the stage. Other than that, awesome location.
I think the prizes were really good. There were the main prizes from the event, some prizes from API Partners and special prizes from some organizations. My team won the Gini API prize and the prize from Friedrich Ebert Foundation.
Food and Drink
It is important to keep the participant hydrated (it is summer after all), and awake. So, a lot soft drinks and caffeinated beverages were provided. There was plenty of food (although not so much variety, specially for those poor souls that do not enjoy pork/bacon, yeah I am talking about you, Karim).
I think the overall organization was good. Natalia Karbasova and the team from Hubert Burda Media made a great job organizing and managing such a large event. After all, there were around 25 teams participating at the same time.
Not everything can be perfect right? I had some issues with how the participants were evaluated. It looked to me that there were two type of contestants. Some trying to write legit hacks and other more worried about producing cool business ideas.
To understand where I am coming from let’s take a look at the definition of Hack and Hackathon based on the always reliable Wikipedia.
Hack is a workaround or quick-and-dirty solution that is clumsy, inelegant, difficult to extend and hard to maintain, yet an effective and quick solution to a problem.
A hackathon (also known as a hack day, hackfest or codefest) is an event in which computer programmers and others involved in software development and hardware development, including graphic designers, interface designers and project managers, collaborate intensively on software projects.
So a Hackathon should produce several quick solutions of existing problems based on software (and sometimes hardware). Not ideas, not pitches, just cool hacks.
In the official description page the guys from Burda Media define the evaluation criteria
- Idea / concept
- Presentation / Documentation
This is already problematic as no particular weight is given to any of those. It also forces teams to spend more time in creating a “pitch” than an actual hack. In my opinion both the actual code (it the sense that it actually does something) and the demo should be the main thing.
Some of the winners (and not winners too) offered great business ideas but they were far away from being cool hacks. Given the flexible interpretation of the requirements, the event ended up comparing Business Ideas vs real hacks, which in my opinion is not the idea of the event (I might be wrong). That is the equivalent of comparing organges with apples.
Tu put it a bit differently, some presentation looked more like they were designed for a pitch competition of a Idea / Startup event, while other where less beautiful but showed real working prototypes.
Analysis of the Winner Project
Let’s take for instance the winners, ShareX Flights, I really dig the idea. The presentation was awesome (check the video below, you will probably agree).
So, from presentation, idea and pitch I would give them a full note. But let’s take a look at the actual hack. Given that all the submission required a Github repository to be set, and a short description to be written, it is easy to check what they did.
The participants reported using the following Frameworks:
Here some technical issues they mentioned:
Bugs in the Gini API → solved by consulting API contact persons
Connection issues with JDBC connector to HANA DB (HANA Tunnel) → We implemented a temporary in-memory buffer
Creating a realistic price curve simulation → Combined randomized periodic curve algorithms
After checking the presentation and the source code, I found they use Gini API to check one file from the email ( via the GMAIL API, with some hardcoded default to make it work, which is fine for a hackathon) and displayed “what they could have saved” by using the proposed app. So the actual connection to Gini API is not part of the Hack but of the landing page of the possible product.
The price data was simulated (as they stated on the description of the project). A sweet work of animation, but not really a hack. In other words, they presented a nice, well designed landing page and a concept of a mobile app. The connection to one of the parters API is not for the hack but for the landing page. So basically the hack did not connect to any external API nor actually did some real work. It was an awesomely crafted simulation. So the question is, where or what is the hack in this work then?
Am I saying that they shouldn’t have won? No, that was up to the judges, they probably had different criteria. However, If I was one of them I would have somehow given more weight to the actual working hack. Because that’s what a hackathon is all about. For a Startup/Idea/Concept competition however, the presentation and simulation would have been amazing.
What could be improved?
I discussed this issues with some people during the hackathon. Many of them agreed with me. Others blamed the sponsor (HvB) as the cause of being less technical. I believe the next event could be improved by:
- More emphasis on the technical part.
- Clearer evaluation metrics.
- No “pitches” or “presentations” or videos, just the demo.
- If you are developing a mobile app, don’t use simulators. run your app directly on the phone.
- Have a technical committee / team to evaluate the source code in private, with the teams.
- Require a live integration, relevant to the hack, with at least at one external API / Data Source (that’s why the partners are for) or another data source.
- Alternatively, two prizes could be offered, such as best technical hack and best business idea could be offered so the two types of project don’t compete for the main prize.
I really liked the event. I think it was the opportunity for learning technologies that I usually don’t have to use in my day to day work. I had the change to meet awesome people ranging from “coders” to designers. I also realized how important a designer can be in such competition, so is almost a requirement to have one in your team.
I am actually looking forward to the next event. Hopefully some of the feedback will be taken into account. If not, I will definitely have designer in my team and prepare myself for Startup pitch rather than a true hackathon.