We Built a Real-Time AI Research Collaborator Into our JOT writing tool | by Vektor Memory | May, 2026 | MediumSitemapOpen in appSign up<br>Sign in

Medium Logo

Get app<br>Write

Search

Sign up<br>Sign in

We Built a Real-Time AI Research Collaborator Into our JOT writing tool

Vektor Memory

7 min read·<br>10 hours ago

Listen

Share

Press enter or click to view image in full size

There’s a moment every technical writer knows. You’re deep in a paragraph about memory migration frameworks or AI transparency, and you realize you’ve been trying to write, but your ideas need further refining.<br>You need a kickstart to find the right direction, and you need it quickly before the moment lapses.

No one’s pushing back on your assumptions. No one’s telling you the paper that directly contradicts your third point or provides you with insights outside of your current thoughts.<br>That’s the gap we decided to close. Over one sprint, we built JOT Collab , a live AI research collaborator embedded directly into the writing interface of VEKTOR Memory’s note editor.<br>Here’s what we built, and some of the challenges we faced.<br>Why Not Just Use Notion, Obsidian, or ChatGPT?<br>This is the obvious question. Here’s the honest answer:<br>Notion and Obsidian are excellent organisers. They store your notes, link your ideas, and keep your writing structured. But they are passive. They don’t read what you’re writing and say “you’re missing something.” They don’t surface the academic paper that directly challenges your third paragraph. They wait for you to ask.<br>ChatGPT and Claude are powerful, but they require you to context-switch. You copy your draft, open a new tab, paste it in, ask a question, get an answer, switch back, wait, collate the data, etc.<br>Every cycle breaks your flow. And they have no deep, persistent magma memory of what you wrote last week, or what insights you’ve already had on this topic.<br>JOT Collab is different in three specific ways:<br>It fires without being asked. Four seconds after you stop typing, it reads what you wrote and surfaces an insight, a suggestion, and four relevant papers — without you leaving the editor or switching context.<br>It builds on your history. Every insight is stored in VEKTOR’s memory graph. Start a new writing session on the same topic and it surfaces what you noticed last time. Your thinking compounds instead of resetting.<br>It challenges your synthesis, not just your text. The insight prompt reads your synthesis section headings and says things like “Section 2 assumes X, but this paper from 2024 argues the opposite.” That’s a different class of feedback from grammar suggestions or general summaries.<br>The closest analogy is a research assistant who has read everything you’ve written, knows the relevant literature, and interrupts you at exactly the right moment, without waiting to be asked.

The Idea: A Collaborator, Not a Chatbot<br>The distinction matters. A chatbot waits to be asked. A collaborator reads over your shoulder and says something when it has something worth saying.<br>JOT Collab watches the THOUGHTS pane. When you pause for four seconds, it fires three things simultaneously:<br>An insight — one sharp observation about what you just wrote, often referencing specific sections of the synthesis pane<br>A suggestion — a declarative sentence you should add but haven’t<br>arXiv papers — four relevant academic papers that arrive via Server-Sent Events after the insight renders<br>The goal was sub-second time-to-insight. The reality was a two-hour debugging session.<br>Press enter or click to view image in full size

This is being tested on Groq llama-3.3–70b-versatile , a fast mid-sized model. And the best part, it is still currently free via API and runs locally.<br>We all are living through a frontier-model greedy token crisis; it's good to see some companies are still offering generous limits via API.<br>The Architecture<br>The system has three layers:<br>Server patch (jot-collab-server-patch.js) — a drop-in Node.js route handler with six endpoints: /api/jot/stream (SSE connection), /api/jot/think (insight + arXiv), /api/jot/suggest (gap suggestions), /api/jot/deepdive (on-demand paper synthesis), /api/jot/article (Medium article builder), and /api/jot/arxiv (proxy).<br>UI layer (jot-collab-ui.js) — 1,000 lines of vanilla JS that injects a collab panel into the synthesis pane, manages SSE connection lifecycle, handles session state (insight accumulation, paper caching, cross-session recall), and wires all the UI actions.<br>Card actions (jot-card-actions.js) — selection toolbar and per-card footer buttons (copy, fix, expand, simplify, summarise, flashcards, → jot) for the DESK chat view.<br>The insight prompt took fifteen iterations. The final version sends: your text, the synthesis section headings, and the last two insights you’ve already seen, so it never repeats the same angle twice.

What needed refining, and Why<br>Press enter or click to view image in full size

The arXiv Race Condition<br>The hardest problem wasn’t the AI calls. It was...