TechNova Hackathon
Challenge
The team that performs the best overall in both the Hack and Pitch categories. They have created a unique and innovative Hack, which encompasses the values presented by TechNova. All teams who submit a project, will be considered for this prize!
This is what created SafeGuard.
As teens on the internet, we are well aware of the effects of social media on mental health. Social media consistently promotes triggering content, such as mental health disorders, eating disorders and serious topics like suicide, to young viewers. A report from the Center for Countering Digital Hate found that vulnerable teenagers’ accounts were sent 12 times as many self-harm videos as standard accounts (The Times, 2022). Young viewers and minority groups who are being profited off of with harmful content deserve a shield between them and the exploitative digital landscape, ensuring their online experiences are safe and empowering.
The Problem
SafeGuard is based around a point system. User can climb up levels by using our features and practicing cyber safety habits.
Let's talk features
⭐ Triggering content detector
Our program possesses the capability to analyze a wide range of multimedia content, including videos and images, and determines if they need to be flagged due to potential triggering or sensitive material. This is our way of standing between our users and harmful or negative content they may be exposed to.
⭐ Impersonation detector
Users may upload and image of themselves for our program to search the web for similar or identical photo's, after which we provide links where those photos are used. This allows those with public, or even private accounts, to protect themselves from impersonation or doxing.
The large majority of the technology we used for SafeGuard we went into cold, with no experience.
For the trigger content detector, we used machine learning to build an AI model which we trained to recognize triggering words and sentence, along with positives ones. We then added features to feed media content to our model. This includes uploading images, then recognizing any text on the image and turning it into a string, and uploading video's to be turned into string-formatted transcripts, which can then be fed to our model to flag for negative words.
The Impersonation Detector uses a SerpAPI in conjunction with Google's reverse-image module to search the internet with an image and return links including the image. Our original attempt at this feature used DeepFace to detect identical faces in images, and we later had to scrap that idea and used Google's image search feature to our advantage.
Figma is a prototyping and design tool, which we used to map out the different screens and functions for our mobile application.
The Problem
When our team went to Hack the North, we were super excited to catch some of the iconic goose merch handed out at the opening ceremony. We sat in what we thought was the best spot to catch the flying geese, right in the front row of the second floor. Unfortunately, every goose launched towards our section either missed or fell down, leaving us empty-handed. Disappointed but inspired, we decided to turn that moment into something fun.
The Solution
We created Geese Love Merch, a game based on our experience at Hack the North. The game takes place in Lazarus Hall (where it all happened!) and stars an awesome goose on wheels as the main character. The goal of the game is simple: fire merch at sponsors who randomly pop up in the audience every 6 seconds. For each sponsor you hit, you score points and get closer to winning the ultimate prize, a dream internship!
How It Works:
Main Character: The goose on wheels with a merch cannon.
Gameplay: Shoot merch at sponsors that randomly appear in different seats.
Timing: Sponsors show up every 6 seconds, keeping the game fast and challenging.
Objective: Hit as many sponsors as possible and rack up points to win.
How We Built It:
We had little to no experience with Unity and C#, but we decided to take on the challenge and learn on the spot. We used Blender to create our 3D models, added animations and sound effects, and gave the game a fun, lively atmosphere. Tools like the Unity Asset Store helped us speed things up, and we used VSCode and GitHub to collaborate.
Challenges We Faced:
Learning Unity & C# in a Day:
We had to teach ourselves how to use Unity and C# within 36 hours. Collaboration Problems: Unity projects can be tricky to work on as a team, pushing changes at the same time caused crashes, which slowed us down. Idea Overload: We had around 4–5 different ideas before finally committing to Geese Love Merch.
IDEATION
Exploring key features to address the complexities of the college experience
Upon taking our user research and noting key pain points to address, we brainstormed and mapped out the preliminary information architecture
through which our product would be structured.
You Made to the Bottom !!!!
While you're here, let's connect:)))
You Made to the Bottom !!!!
While you're here, let's connect:)))
Email : samriddhi.makasare@gmail.com
Email : samriddhi.makasare@gmail.com
Linkedin :https://www.linkedin.com/in/samriddhimakasare/
Linkedin :https://www.linkedin.com/in/samriddhimakasare/
Linkedin : https://www.linkedin.com/in/samriddhimakasare/
Made with lots of love and espresso
SAMRIDDHI MAKASARE | 2025