See trends before they go mainstream.
eonik tracks how ideas spread across the internet and surfaces emerging formats, memes, and content ideas you can post before the trend peaks.
Congratulations on the launch!
I tried a famous person's twitter handle and the app just froze.
Also there's no "about" page - it would be helpful to know the class of algorithms or training data you used, because otherwise we'll assume you're just using a generic large language model (which doesn't feel well suited to statistical trend analysis, especially when there's low training data like with new products or trends)
Report
Maker
@hex_miller_bakewell Thank you very much for trying it out and sorry for the fail! It is the first version so we have only kept the instagram handle as input source. We will add multi platform input support ASAP. For the language model, we are directly sourcing data from the sources using APIs and scrapers and RAG based grounded searches with LLMs.
The suggestion for new post and the analysis of the fetched posts is currently done by language models.
Point taken! We will add details about the process to build the authenticity in the trend detection mechanism.
Replies
Hey Hunters & Builders! 👋
I'm Abinash, co-founder of eonik.
I’ve spent years watching creators and brands struggle with the same question every day:
“What should we post next?”
Most tools show trends after they explode.
But internet culture doesn’t work randomly.
Ideas propagate through predictable ecosystems: Reddit, Twitter, creator communities, then TikTok and Instagram.
We built Trending Now to detect those signals early.
Instead of showing you yesterday’s trends, it tries to answer:
What will trend next?
Still early and would love feedback from creators, marketers and builders here.
Would love to know:
👉 Where do you currently discover trends?
👉 What signals help you decide what to post?
Thanks for checking it out 🙏
Interesting
@madalina_barbu Thank you very much!
@hex_miller_bakewell Thank you very much for trying it out and sorry for the fail! It is the first version so we have only kept the instagram handle as input source. We will add multi platform input support ASAP. For the language model, we are directly sourcing data from the sources using APIs and scrapers and RAG based grounded searches with LLMs.
The suggestion for new post and the analysis of the fetched posts is currently done by language models.
Point taken! We will add details about the process to build the authenticity in the trend detection mechanism.