Last December, we and six other Skolkovo companies conducted our own startup hackathon. Without corporate sponsors and any external support, through the efforts of the programming community, we gathered two hundred participants from 20 cities of Russia. Below I will tell you how we did it, what pitfalls we met along the way, and why we immediately began to work with one of the winning teams.
The interface of the application that controls the Watts Battery modules from the finalists of the track, “Wet Hair”
Company
Our company Watts Battery creates modular portable power plants. The product is a portable power station 46x36x11 cm, capable of producing from 1.5 to 15 kilowatts per hour. Four of these modules can provide power to a small country house for two days.
Although last year we started shipping serial models, in all respects the Watts Battery is a startup. The company was founded in 2016 and since that year is a resident of the Skolkovo Energy Efficiency Technology Cluster, today we have 15 employees and a huge backlog of what we would like to do at some stage, but not right now.
It includes purely software tasks. Why?
– . , , . , , .
– , , , , , . , , ( ), , .
, , – . . , -.
, ,
: . .
, « », : - , , . . , , .
: , , , . , 48 , , , , . , .
- , , , .
. , , - , , , , «» , . , , , , , . .
, . , , , , : , .. - .
Watts Battery , , ( ). , . , , ( ).
, : : , , . : agile , ( ).
, Amazon AWS ( , Amazon , , ). Amazon- . , - . - , - . , Nubble ., - . ( ), .
, . , .
, – «», () .
: , . , , .
: . (, ). , , – . – . : , – . . , Watts Battery , , . – , – , . , , . , , . - , , , (, , ). , , , , . - , . , . , 500 , – 3,5 – ! : « , -».
opensource Yandex CatBoost
: , , ( , – , ), , , , , . , -, , . , , , . , : - , , , - .
: – . Watts Battery, , . – , Watts Battery . machine learning-, – , , – . . , , - .
: . 48 , , , . , .
: ( «» , , ), Watts Battery, . , feasibility , , , , , .
: , , . , .
opensource Facebook Prophet
: , roadmap, . , , , , . , .
: () Watts Battery, , , .
: - , , , . , , smart . , . , , , .
: , -, .
: .
: , , -. , , , - – – Arduino ( ), : , .
* Speech by the winners of the second track *What they got as a result : two teams offered full-fledged business ideas for their work: one focused more on the Russian segment and the other on the foreign one. That is, in the final they didn’t just tell how they came up with the application, but in fact they came to do business around Watts. The guys outlined how they see the use of Watts in several business models, cited statistics, showed in which regions which problems are adopted, where laws are passed, outlined the global trend: it’s unfair to mine bitcoins, fashionable to mine kilowatts. They purposefully came to alternative energy, which we really liked. The fact that the participants, in addition, were able to make a still working technical solution, suggests that they can independently launch a startup.Main conclusion: There are teams ready to take Watts Battery as the basis of their business model, develop it, become partners / partners of the company. Some of them even know how to highlight MVP business ideas and work on it in the first place, which is currently lacking everywhere in the industry. People do not understand when to stop, when they need to bring to market a solution, albeit an early one, but one that works. In fact, the stage of polishing a solution often does not end, the technical solution passes the line of reasonable complexity, enters the market overloaded, it is not clear what the idea was originally there, what targeting customers, what business models are in place. Like in a joke about Akunin, who wrote another book when he signed the previous one. And here it was done in its pure form: here is a graph, here is a counter, here are indicators, here is a prediction - that's it,nothing else is needed to run. With this, you can go to the investor and get money to start a business. Those who found this balance, and left the track winners.Advice for the future : at the next hackathon (we are planning it in March this year ), it may make sense to experiment with hardware. We have our own hardware development (one of the advantages of Watts), we completely control the production and testing of everything we do, but there are not enough resources to test some "iron" hypotheses. It may well be that in the community of system and low-level programmers and hardware developers there are those who will help us with this and in the future will become our partner in this direction.People
At the hackathon, we expected more of those who want to try themselves in a new field (for example, graduates of various programming schools) than those who specialize in such developments. But still, we expected that they would do some preparatory work before the hackathon, read how energy consumption is generally predicted and how the Internet of things systems are structured. So that everyone comes not just for fun for interesting data and tasks, but also with a preliminary immersion in the subject area. For our part, we understand that for this it is necessary to publish in advance the available data, their description and more precise requirements for the result, publish API modules, etc.The technological level at all was approximately the same, plus or minus opportunities, too. Against this background, not the last factor was the level of harmony. A number of teams did not fire, because they could not clearly divide by areas of work. There were those in which one person did the entire development, the rest were involved in preparing the presentation, in others - someone was given tasks that they probably did for the first time in their lives.Mostly the participants were young, this does not mean that among them there were no strong machine learning engineers and developers. Most came in teams, there were practically no loners. Everyone wanted to win, someone wanted to find a job in the future, about 20% have already found it, I think this figure will grow.We did not have enough hardware geeks, but we hope to compensate for this at the second hackathon.Hackathon move
As I wrote above, most of the 48 hours of the hackathon we were with the participants and, following their success at checkpoints, we tried to adapt the assignment and acceptance conditions of the first, analytical track so that, on the one hand, the participants could finish it for the remaining time, and on the other - it was of interest to us.The last refinement to the task was done somewhere in the region of the last checkpoint, on Saturday afternoon (the final was scheduled for Sunday evening). We simplified things a bit more: removed the requirement to convert the model to new data, left the data that the teams already worked with. Comparison of the metrics gave us nothing, they already had ready-made results according to the available data, and on the second day the guys were already tired. Therefore, we decided to torment them less.However, three out of four participants did not reach the final. One team already at the start realized that they were more interested in the track of our colleagues, the other just before the final realized that in the process of processing it had filtered the necessary data ahead of time and refused to submit its work.The team “21 (Effect of wet hair)” until the last participated in both of our tracks. They wanted to cover everything at once: machine learning, development, application, and website. Until we threatened them with withdrawal at the last moment, they believed that they were doing everything, although it was already obvious at the second checkpoint that the main thing - machine learning - could not achieve significant progress: they generally managed the second block, but predicted electricity consumption were not ready. As a result, when we determined the minimum task for qualification for the first, they still made a choice in favor of the second track.Fit-predict had a balanced roster tailored for data analytics, so they were able to overcome everything. It was noticeable that the guys were interested in “feeling” the real industrial data. They immediately concentrated on the main thing: they analyzed, cleaned the data, and dealt with each anomaly. The fact that they were able to build a working model during the hackathon period is a great achievement. In working practice, this usually takes weeks: while the data is being cleaned, while they are digging into it. Therefore, we will definitely work with them.In the second track (management), we expected everyone to do everything in half a day and come to ask to complicate the task. In practice, they barely managed to complete the basic task. We worked on JS, Python, which reflects the current state of the industry.Here, the worked teams have also achieved results, in which the division of labor has been built, it is clear who does what.The third team, FSociety, seemed to have a solution, but in the end they decided not to show their development, they said that they did not consider it working. We respect this and did not argue.The team “Strippers from Baku” won, which was able to stop itself, not to chase the “little things”, but to make an MVP, which is not a shame to show and which is clear that it is possible to develop and scale further. We immediately told them: that the additional features are not of much interest to us. If they want to register through a QR code, face recognition, let them first make graphics in the application, and then take on the optional.In this track, “Wet Hair” confidently entered the final, and with them and “Strippers” we discussed further cooperation. We have already met with the latter in the new year.I hope that everything works out, and we are waiting for everyone at the second hackathon in March!