Lattics

Brain-like knowledge base with AI writing & deep research

Productivity

Writing

Artificial Intelligence

Visit WebsiteSee on Product HuntMicrosoft ⧉App Store ⧉

curious how the AI writing layer plays with the brain-like graph structure over time - does it auto-link a new AI-drafted note into the right place in your existing knowledge graph, or do you still have to manually connect it after generating? that manual reconnection step is usually where these tools quietly turn back into a flat notes app once you've got a few hundred entries.

the way lattics weaves AI directly into the research flow instead of bolting it on feels really thoughtful, like the team actually understood how writers get stuck before shipping anything. excited to dig into this one.

The "brain-like" framing is interesting — most knowledge-base tools are still fundamentally folders and tags with search bolted on. Curious how the underlying structure actually works: is it building real semantic connections between notes over time (more like a knowledge graph), or is "brain-like" more about the retrieval/writing experience on top of a fairly standard store? That distinction matters a lot for how well it holds up as someone's knowledge base grows into the thousands of notes.

Hey Louis, that scattered feeling of jumping between a dozen windows is exactly what wrecks my focus when I settle into a long piece of writing. Bringing calm to that mess really appeals to me, and knowing my work stays on my own computer is a lovely bonus.

Brain like knowledge base is a bold claim but the concept of non-linear note organization is genuinely useful. How does the AI writing side integrate with the knowledge graph, does it pull context automaticaly?

The way Lattics pulls my scattered notes into a visual map while I draft has genuinely changed how I approach long projects. The AI suggestions feel tuned to what I'm actually working on, not generic fluff.

The local-first angle is the strongest part here. For research and thesis writing, having cards, timelines, citations, PDFs, and drafts in one desktop workspace sounds much better than bouncing between a notes app, reference manager, and chat window.

The main thing I’d want to understand is how the AI layer handles privacy. When using right-click rewrite or @ mentions across multiple articles, is that always sent to a cloud model, or can users choose what leaves the machine?

I noticed the PDF translation feature. Does the translated version stay aligned with the original document layout? And can it also translate text inside images or scanned pages?

the privacy-first local storage plus AI writing combo is the part I'd want more detail on - the moment I right-click and ask for a rewrite or start an AI chat, that selection has to leave the machine to reach whatever model is doing the work, right? curious whether that's opt-in per action, and whether there's ever a local/offline model option for people who picked this specifically for the local-storage privacy angle in the first place.