This past weekend, AssemblyAI was a proud sponsor of DeltaHacks, an annual student-run hackathon at McMaster University. We saw a great turnout at the event, with over 300 participants and 95+ projects built in just under 36 hours!
As a DeltaHacks sponsor, we offered our Speech-to-Text API as well as our Audio Intelligence APIs as resources for students to hack with and provided support for our asynchronous transcription API, real-time streaming API, as well as our Audio Intelligence APIs like Chapter Summarization and Sentiment Analysis. We also had the difficult job of choosing the best project that used the AssemblyAI API.
Here’s a highlight of the top three projects built using AssemblyAI Speech-to-Text APIs:
An accessible environment for jobseekers to practice and improve their interview skills.
What Does This App Do?
Interviewy is an application that allows users to practice their interviewing skills by analyzing and providing feedback on their video recorded interviews. This interface was built using the MERN stack in the backend. The AssemblyAI APIs are used to detect the sentiment and key topics covered in the interview response.
Users can rewatch their recorded interview and review their performance based on sentiment as well as topics covered, corresponding to a specific time-stamp in the report.
Inspiration for the App?
Even qualified applicants can be rejected for jobs simply because they were unable to effectively showcase their knowledge and skills during the interview. Since the start of the pandemic, the virtual interviewing format has been challenging for many applicants, especially students, as interviewees have reduced access to in-person resources where they can develop their interview skills.
Improve speech and public speaking skills by analyzing sentiment, filler words, and speaking speed in your speech.
What Does This App Do?
UmmLess.tech helps users to improve their public speaking abilities by providing feedback on their recorded speech. Using the AssemblyAI APIs, the application allows users to upload audio files, which are analyzed for sentiment, filler words, and speaking speed. This analysis is then presented to the user in a user-friendly, easy-to-read format, so they can identify areas of improvement in their next presentation.
Inspiration for the App?
Approximately 77% of the US population feel some anxiety with public speaking. Commonly, this causes people to speak faster, stutter, or speak with an abundance of “ums” or “uhs”. Particularly during the pandemic, where virtual presentations have become the new norm, it can be hard to read the audience or get audience feedback.
An education and productivity platform that creates dynamic study sheets to increase active learning based on lectures or videos in real time.
What Does This App Do?
Annotate.io is an education and productivity application that allows users to automatically generate notes based on lecture content. Users can submit a .mp4 file or a YouTube link and get interactive notes. Using the AssemblyAI APIs, Annotate.io is able to detect topics, summarize content and highlight key topics in the lecture. Using Twilio, users can then email a pdf version of their notes to themselves and also share their notes to others.
Inspiration for the App?
With online lectures becoming the standard way of learning, it can be hard to consistently make clear and concise notes. What if there was an application that could transcribe your online lectures and summarize the key concepts so you don’t need to go through hours of lectures to find one key note? Annotate.io is a note-taking assistant that does just that. With Annotate.io, you can build concise study sheets based on your lectures.
We are highly impressed with the work that went into all the projects, and honored to have been able to sponsor so many creative use-cases for the AssemblyAI Audio Intelligence APIs! We can’t wait for another round of DeltaHacks next year!
