Designing SuperBrain's Knowledge Retrieval System by Rod AcevedoDesigning SuperBrain's Knowledge Retrieval System by Rod Acevedo

Designing SuperBrain's Knowledge Retrieval System

Rod Acevedo

Rod Acevedo

SuperBrain

Most of what we consume disappears. SuperBrain captures it, connects it, and gives it back to you at the moment it's useful.

You've already solved most of the problems you're facing. You read the article, heard the podcast, had the conversation. But the insight is gone, because your brain wasn't designed to index thousands of inputs across months of consumption.
The cognitive load isn't in learning new things. It's in failing to retrieve what you already know at the moment you need it. Every bookmarked article you never revisit, every podcast that faded by the next morning: that's lost value, because nothing connected it back to the challenge in front of you right now.

Impact

SuperBrain removes that retrieval burden. It captures what you read, listen to and say through three input modes, consolidates everything into a journaling layer, and surfaces the right pieces when a new challenge maps to something you've already encountered. You stop carrying the weight of remembering where you learned something and start operating with the full benefit of your accumulated experience.

My role as Founder & Product Design engineer

The product vision started with one insight: the value isn't in any single input, it's in the connections between them over time. So rather than building three separate tools, I designed a unified system where Read (articles, books, highlights), Listen (podcasts, voice memos) and Speak (conversational journaling) all feed the same knowledge layer. The journaling sits on top as the reflection and retrieval interface, connecting today's challenge to the article you read three months ago and the reflection where you worked through something similar. Patterns in your own thinking become visible.
I owned everything: product vision, MVP scope, interaction model, agent architecture, prompt engineering, and iteration based on real usage. The key product decisions were around how much the agent should proactively surface versus wait to be asked, and how to make retrieval feel like a conversation rather than a search query.
Interested in the story behind this project? Let's have a chat!
Like this project

Posted Jun 10, 2026

Most of what we consume disappears. SuperBrain captures it, connects it, and gives it back to you at the moment it's useful.