3 tips on winning a hackathon as a solo contestant.

I’m now on my 4th year doing data science as livelihood! It has been fulfilling so far, but it just occurred to me one time that I have not been participating in community data science activities outside of my work.

Most of my colleagues and students have participated in a hackathon as part of their training and exposure. I havent had that experience yet!

Last month, Cirrolytix announced that they would be conducting a Dataquest hackathon in partnership with UN Women, International Labor Organization, International Organization for Migration and Data Ethics PH. The theme was Gender-Responsive Labor Migration and it aimed to support migration governance agencies through capacity-building and advocacy interventions centered on data through an ethical and moral lens.

This is a great chance to test my data science skills so far and contribute to a noble cause, so I told myself that I should join!

The Output#

For this hackathon, I created the OWWAksyon smart chatbot ^* to help OWWA’s Direktang Aksyon livestream program and generated insights with a special analysis on the issues raised by OFW women.

We were asked to create a 5 minute pitch showcasing our solution and data science product. Here is the video of my final pitch.

A panel of 8 expert judges from both the data science and the migration and gender fields assessed the final pitches―probably the most intense scrutiny on my work since my master’s thesis defense.

Among 10 finalists, this pitch was awarded the Most Practical Solution, the solution with the most plausible concept and design that is most ready to be adapted by a migration governance agency.

Working solo in a hackathon#

It had also occurred to me that I was the only solo contestant who made it to the pitch finals. This is an unusual arrangement for participating in hackathons, and it has its advantages and disadvantages.

Disadvantages#

I’ll list the disadvantages first!

1. You bear all the workload. You will only have a single brain and a pair of hands to do all the work required for your output. If the hackathon asks to form a 3-4 person group for team participants, we could expect a solo participant to handle a 3-4 person workload.
2. You will have less diverse ideas. You will have no one to throw around ideas with during the proposal stage so you will have to compensate with longer research and scoping hours for your planned solution.
3. You will miss out on networking opportunities. A hackathon is also a chance for others to get to know your work quality, ethics, and coordination skills, as well as your personality. Although you can network during the hackathon events, that time is short compared to the time spent with the team working on the output. You are also less likely to form bonds with your co-participants.

Advantages#

But there definitely are advantages!

1. You work at your own time. Groupwork requires lots of coordination to find common time to align with the group. For someone with a fulltime day job and (recently) a night owl like me, my optimal working hours may be too late for potential teammates. I’m glad I can work on my preferred time without inconveniencing anyone.
2. You make all of the decisions. Getting people to agree is hard, especially if you are following a tight timeline. Working solo, I want to do away with this and save time because I have a lot to code and design for the output.
3. You take home the entire prize if you win. Probably the most practical one! You will finish the output on your own, and if ever you win, you dont have to share the prize.

Tips on winning as a solo contestant#

It was my first time to join a hackathon and I’m very happy that my work has passed the scrutiny of the judges and qualified for an award.

For those who want to tread the same path, I have a few tips here that worked best for me and might work for you too!

1. Leverage knowledge and code you are already familiar with.

   I cannot emphasize enough that there is much less time that you can dedicate on a project working solo compared to a team. There are a myriad of possible solutions you can pitch and prototype, but given your limitations, you will have to select the one which you are most confident that you can build within the deadline.

   It is my first time to work with a gender migration use case, and familiarizing with the theme is taking a lot of my time. I had to figure out an angle that would best serve the target stakeholders while maximizing my current skill set.

   After long hours of researching, I stumbled upon OWWA’s Direktang Aksyon facebook livestream program and saw some weaknesses that needs a data system intervention. I would not have finished it in time have I not done previous work on public Facebook page data. As much as possible, reuse previous notebooks and scripts!

2. Act on the data fast! Not long after I had decided on my output, I was hit by the scale of the data that I will have to take in to show robust insights and to get my prototype working. Each Direktang Aksyon livestream was gaining around 2-3k comments, and I stated that I would be covering 15 of these livestreams.

   This was the only project who took on data on the 20000+ rows range and I knew I had to step it up, fast! I had to prepare a custom NLP stopwords corpus and lookup method for Taglish comments with spelling quirks. I also had to whip up a gender guesser model using Facebook usernames. It was very tedious but I know the insights I’ll obtain will best serve our OFWs and OWWA admins.

   All in all, data exploration and cleaning took up 70% of the time. As soon as you have data, work on it!

3. Scope the project to address a very specific need of a very specific beneficiary

   There are a lot of other exciting solutions that could have been feasible if I was working in a group, but as an individual, you have to scope small. This doesnt have to be a disadvantage though, as you can work on a highly-targeted solution.

   To do this, you should work on the hackathon project as if you are the stakeholder. How do I do this?

   • Imagine working for a migration agency–what kinds of solutions can I create in 2 weeks that will help relieve pain points in a specific process?
   • Imagine that you are an OFW seeking assitance–what kinds of solutions can I create in 2 weeks that will assure me that my concern can be heard?

   The smart chatbot format I chose ensures that all OFW concerns are recorded privately and securely (cf. the public comments section) and saves a lot of OWWA admin time filtering, tracing and digesting each of these comments, reallocating their precious time to actual addressing of the problems.

   The key idea here is you should be doing more listening and research to understand the client’s needs than you persuading them to understand and use the solution you built.

A great solution may come jampacked with fancy features, but a winning solution, while less complex, builds stakeholder trust through a deep understanding of the situation and thus will need less convincing to be chosen.

I hope this was useful! Thank you so much and see you in the next blog!

JCP: Took the project demo OWWAksyon Facebook page down because I was getting messages from OFWs thinking that it was an OWWA affiliated page. I also restricted the data backend for privacy. I’m afraid the QR codes in the pitch video do not work anymore, but please do reach out to me personally if you want to learn more about my solution!