Last weekend, AssemblyAI held our first-ever hackathon. With the AssemblyAI $50k Winter Hackathon, we hoped to foster creativity in building AI-first products. During the hackathon, 440 participants from 84 countries worked hard to build over 150 projects.
With so many great projects to choose from, it was hard to narrow them down to the winners. We were incredibly blown away by the quality and quantity of submissions and want to congratulate all of the hackers that came together with us to create such a special event!
Learn more about the event, including highlighted projects and guest judges/mentors, below. You can follow us on Twitter to stay in the loop for our next hackathon!
First Place - Superpaint
Project Description: Augment your drawings with AI! Superpaint is a drawing app which allows users to sketch images and use Stable Diffusion to inpaint regions of the image in a text-guided way seamlessly.
Second Place - Toy Story Creator
Project Description: Children spend the entire day living in a fantasy world they create for their toys. What if we could bring that world to life with a bedtime story as they drift off to sleep?
A user uploads a few images of their toy, and Toy Story Creator generates a suitable story for the toy and illustrations to go with it.
Third Place - OperatorAI
Project Description: OperatorAI is a revolutionary new AI system that helps triage 911 calls and connect callers with emergency services. With OperatorAI, you can trust that your call will be handled quickly and efficiently, even during times of high call volume or natural disasters. Thanks to its advanced technology, OperatorAI can also accurately locate callers, ensuring that help arrives as quickly as possible. So when you need help in an emergency, trust OperatorAI to help you get the assistance you need.
Best Project Built with AssemblyAI - Pupil.ai
Project Description: Pupil is a chrome extension that links to an AI chatbot which answers questions about educational videos. Whenever the student pauses the video, Pupil will ask them if they have any questions. The AI has access to transcript information about the video, so its responses consider the section of the video the student is currently watching.
Best Project Built with Cohere - Docspace
Project Description: Docspace aims to fill a void in the kind of information that is typically available with legal research products by diving into the text of complaints (and in the future judge orders) and identifying the specific claims, counts, and causes of action relevant to a case.
Check out some other awesome projects below - the Honorable Mentions. Note that they are listed in no particular order.
Project Description: Introducing Supercut, your personal autonomous clip maker. Simply upload a podcast video recording, and Supercut will automatically identify parts that would make good clips and create the clips for you. It even generates Tweets for you to post alongside your clips!
Project Description: Using Photostudio, anyone can create stunning product images for real products in less than 30 seconds using only a phone camera. To use Photostudio, a user clicks a picture using their phone and then multiple AI Image algorithms are used to generate stunning product images while preserving all details of the original product.
Project Description: Scaling Potato is a modern solution to help people prepare for technical interviews by practicing with an AI interviewer. Users run through mock interviews in which the user and the AI interviewer can converse to solve a coding problem. The user can also use the built-in code editor we added to write and execute their code for the problem. Questions are crafted in a strategic way and modeled from questions asked in real interviews. At the end of the interview, our platform analyzes the entire practice interview data to crunch out actionable results and give feedback.
Project Description: MaiBook is a personalized learning platform for kids of all ages! MaiBook creates bespoke books with comic illustrations to take children on an adventure that teaches them a target learning topic! Children get to learn via fun content where they can see themselves immersed in an adventure, allowing them to connect with the material and retain the content better.
Project Description: MidJournal uses AI models to transcribe your spoken thoughts into a journal entry and generate a complementary image that captures its essence. Simply log in to access your journal, create a new entry by talking out your reflections, and watch as your audio is transcribed in real-time. Edit your text entry as desired, then submit it to your journal. Your new text entry will appear, and soon after, you'll see a cool image corresponding to the text.
Congratulations to all of the hackers that came together with us to create such a special hackathon.
Narrowing down the winners was tough with so many great projects submitted. Check out some of the other randomly-selected submissions below, or check out the hackathon's Devpost gallery to see all submissions.
A Special Thanks
We'd like to give a special thanks to our guest judges and mentors - this event would not have been possible without you!
- Nat Friedman (investor, former partner of Y Combinator)
- Daniel Gross (investor, former CEO of GitHub)
- Lenny Rachitsky (author, investor, former product lead at Airbnb)
- Omar Sanseviero (machine learning engineer lead at Hugging Face)
- Nick Frosst (cofounder of Cohere)
- Chanin Nantasenamat (Senior Developer Advocate Snowflake)
- Daniel Kornaś (ML Engineer at Infineon Technologies)
- Jaydeep Karale (Software Engineer at UBS)
- Sophia Yang (Senior Data Scientist at Anaconda, Inc.)
- Shubham Saboo (Head of DevRel at Tenstorrent Inc.)
With our first-ever hackathon, we hoped to foster innovation and creativity for AI-first products. The great ideas, energy, and collaboration from all hackers involved helped us realize this to a greater level than we could've hoped, so we want to give one final thank you and congratulations to everyone involved!