Seekedia

• 

  homepage

Inspiration

As an avid podcast listener, I have often thought the podcasting community can benefit from few improvements. With the typical podcast being 1-3 hrs long, I wanted the ability to easily search through the conversations or check if they mentioned particular topics without having to sit through the entire thing. Simply using keyword searches is not enough; I wanted a tool that can understand the context of the discussion and highlight relevant segments like Seekedia. Seekedia is phase one of a two phase project. Phase 2, which is already in development is a platform for the podcasting community with a built-in searching capability of Seekedia. Think twitter for podcast.

What it does

Seekedia is a ai powered search engine for audio/video/podcasts

How we built it

It works by transcribing audio/video podcasts using assemblyai using cohere.io to generate embeddings of similar sentences using pinecone.io to store and search through podcasts

Challenges we ran into

Not enough time as a solo developer

Accomplishments that we're proud of

created a beautiful ui/ux prouct

What we learned

Having found this hackathon at the last minute, I decided to join the competition solo. While it was a strenuous experience, I learned that a lot can be archived in a short amount of time when there is focus, determination and love for the work.

What's next for Seekedia

Seekedia is phase one of a 2 phase project. Phase 2 podcasting social media platform that is currently in development. On this platform, users can search topics from millions of podcast episodes, find all related mentions of said topic as well as related subjects. They also have the ability to upvote, share, like and comment, creating discussions surrounding the topics of the video. From creator perspective, they will benefit from Seekedia’s search engine which unlocks their media content making it more discoverable/google searchable boosting SEO benefits.

Built With

• assemblyai
• cohere
• pincecone