Several years ago, our team developed iSearch, a product from a project aiming to develop a deep learning–powered search engine for biomedical literature. You can still access it today at biokde.insilicom.com.
At that time, we were thrilled by the results.
For complex scientific queries, iSearch consistently outperformed PubMed and Google Scholar, even Semantic Scholar, while strong, still fell short. We thought researchers would naturally adopt it once they saw how powerful it was.
But we quickly learned a humbling lesson:
👉 Great technology isn’t enough.
👉 People rarely change habits without a compelling reason.
👉 Even the best tools can go unnoticed without a clear path to reach users.
We had no real strategy for commercialization or user adoption, and as a result, the project didn’t gain the traction we had hoped for.
Still, I personally use iSearch every day. It remains, in my opinion, the most powerful search engine for biomedical literature. And interestingly, the algorithms behind it later helped us win the BioASQ competition this year (document retrieval in the biomedical question-answering track), just not in the way we had originally envisioned.
This project taught me a lasting lesson:
💡 A product’s success depends as much on how you deliver it to the world as on how well it’s built.
Have you ever created something great that didn’t get the attention it deserved?
I’d love to hear your stories and what you learned from them.
#AI #DeepLearning #BiomedicalResearch #Entrepreneurship #Innovation