When I do real research — reading papers, books, or trying to understand a complex topic — it never happens in one sitting.

I explore a bit, get interrupted, come back days later, ask new questions, find new sources, and gradually build understanding.

Existing AI tools are great at answering questions once, but they reset context every time. Notes tools preserve text, but not understanding — they don’t help me reconnect ideas, revisit sources, or see how my thinking has evolved.

I built Kerns to sit in between.

Kerns is an AI research workspace designed for research that unfolds over time. Each topic lives in a long-lived space where:

• You can ask new questions as they emerge • Upload PDFs, papers, or documents alongside discovered sources • Run deep research in parallel, not one query at a time • Move fluidly between summaries and original source text • Revisit work weeks later without losing context

Instead of scattering work across chats, documents, and bookmarks, research accumulates in one place and compounds as you return to it.

I’ve been using Kerns myself for things like:

• Reading and synthesizing books and essays • Comparing technical papers and model releases • Tracking ongoing policy or industry topics as they evolve

I’d love feedback from anyone who does sustained research — analysts, researchers, founders, students, or serious self-learners.

Thank you!