Google Girl Hackathon: A Surprising Win!!!

Hey there! I am Pakhi Srivastava, a final-year student pursuing a B.Tech in CSE from Sardar Vallabhbhai National Institute of Technology. I have previously interned at Flipkart and Atlassian as an SWE Intern. Recently, I won the Google Girl Hackathon 2023. Through this blog, I will be sharing my experience in detail for the same.

Girl Hackathon is a program for women students in computer science and allied courses across India. It has been designed to provide a platform for women in the engineering campus space to showcase their coding skills and promote creative solutions for real-time technical challenges. This virtual hackathon challenges the innovation and problem-solving skills of the participants and inspires them to creatively engage with Google.

Source: https://rsvp.withgoogle.com/events/girl-hackathon-2023/home

Interested folks had to register on the official website before the deadline. After that, a series of rounds were conducted. Let us have a look at all of those in detail!

Round 1: Google Online Challenge (3 June, 2023)

The first round was a coding round where 2 questions were given based on DSA. There were multiple sets of questions in which a wide range of topics were covered including stacks, binary search, and bit manipulation to name a few. The difficulty level of the questions was LeetCode medium-hard. The time limit was 1 hour to solve both the questions.

Selection criteria: Participants were judged on their performance in the coding challenge alongside their individual profile screening through resumes

Resources: Leetcode

Round 2: Google Learning Cohorts Workshops (17 June to 24 June, 2023)

All the selected participants from Round 1 were divided into different learning cohorts according to their performance in previous rounds. I was put into the following workshops:

Workshop 1: Art of writing Idiomatic Code

In this workshop the primary focus was on teaching the participants the practice of writing idiomatic codes in the language of their choice & how it makes the process of understanding & debugging easier for all the people working on any codebase (or project) alongside us. It also taught me a rather important thing: write comments to explain the unexplainable behavior of a code.

Workshop 2: Pattern Recognition in DSA

This workshop focused on recognizing what kind of patterns are present in a problem & what algorithms & data structures can be used to solve it. It also included the optimization of the code to reduce time complexity & space complexity.

These workshops were followed by a DSA coding round where the participants had to solve 2 questions based on varied topics such as divide & conquer etc. The time limit to solve both questions was 1 hour & the difficulty was LeetCode medium-hard.

Round 3: Ideathon Round (4 July, 2023 )

After the second round, the hackathon entered its pre-final phase & this is where I feel the competition intensified. The primary focus was on developing solutions to growing environmental challenges using AI. The primary problem statement was further subdivided into 3 themes. The participants were required to choose one of them & develop a solution for it.

The submission was to be made in the form of a Google Doc, in a predefined template. The process, problem, solution as well as all the strategies to be used for the expansion & deployment of the solution so that it can be used effectively in real-life scenarios was required to be explained in full detail in the submission doc. Along with the doc, the link to the GitHub repository was required to be submitted as well which contained the code for the solution, as well as a link to the README file. The time limit was approx. 5 days more or less.

Approach
First & foremost I surfed through the net to check out the resources that I can gather for each problem statement along with rough assessment of the impact of the solution that I was going to propose. Once I had settled down on a problem statement & had gathered resources, I went through some of the research papers which were on a problem statement similar to the one chosen by mine to checkout the process & methods they had employed as well to get the assessment criteria for the selection & performance of a model. After the initial steps of pre-processing data, I took 3–4 models & tested each of them to check which one fit the best. This process helped me clearly establish what I wanted to achieve rather than jumping headfirst into coding.

While documenting I took into account the evaluation rubrics:

1. Potential Impact of Proposed Idea
2. Usage of correct algorithm and AI technique
3. Code Quality
4. Appropriateness of Data Structure
5. Testing

The Grand Finale: Presentation round (14 July, 2023 )

To my surprise 😮 my idea got accepted & I was one of the 45 finalists who had made it to the finale out of over 26,000 applicants. For the final presentation, the participants had to create a presentation on Google Slides showcasing their idea along with a demo of the working prototype. In the final 4 days before the grand finale I worked on making some tweaks to my already existing solution & added a part of the future scope to my working prototype.

After all the tweaks & additions were done I worked on the presentation. The template for the ppt was given by the organizing team. I made sure to include the diagrammatic representation of my idea & kept the presentation as simple as possible. I took care to explain my idea, it’s working, along with impact & future scope in as few words & more pictures & diagrams as possible. I also took care to craft my presentation according to the evaluation rubrics given by them:

1. Presentation and Communication
2. Technical Implementation
3. Impact and Potential

Along with this I also prepared a small writeup for the presentation.

On the day of the finale, all the finalists were divided into groups & there were 3 tracks. The presentations were being presented on the screen by the Girl Hackathon team itself. A total of 12 minutes was given for the presentation of the presentation + demo & 4 minutes for questions. The judges were quite thorough in asking the questions. For my presentation, most of the questions were centered around the future scope & expansion of the current project. To my (yet another!!!!) surprise I was crowned as the winner of the hackathon (YAYYY!!! 🔥).

This was indeed an experience of a lifetime. I never imagined when I had applied for the hackathon that in a month’s time, I would be the one taking the trophy home. Through this hackathon, I have become more comfortable with stepping out of my comfort zone (since I was a web & app developer, & had done AI/ML mainly as a part of my college course). I have also learned that the path to victory is surely never easy but the vigor to never give up helps a lot in more ways than we can imagine.

The uneasiness of not knowing where to start & then keep building on the original idea through trial & error is also something that I have learned in the process of GGH. The thrill of competition & the urge to build something unique that can actually help people in real life motivated me to keep hustling.

I also learned how to properly do research, the importance of documenting stuff & most importantly data preprocessing & representation (hell of a task, my salute to all data scientists out there 🖖). Apart from this while searching for a solution I got to develop multiple models & in the process, I learned more about their underlying algorithms.

Tips I would like to give to future aspirants

1. Keep practicing DSA and give contests to get a stronghold over it.
2. Do not limit yourself to a particular stack etc. Develop an attitude to solve problems through whatever means necessary (in short, love solving problems & not the tech stack). This helps a lot in adjusting in the long run.
3. Embrace the challenges and most importantly cherish the journey — you may win, you may not but the knowledge that you have gained along the way is there to stay.

In case you need any help, feel free to reach out to me on LinkedIn. If you find this article helpful, hit the clap button and follow Codess.Cafe for more such articles.

Codess.Cafe: LinkedIn | YouTube | Website | Medium | Twitter