Your AI Keeps Forgetting Your Research. Heres How to Fix It
Transcript
[0:03] Why Do AI Models Forget Research Across Different Chat Sessions and Platforms?
Speaker: Karan Bavandi
Karan Bavandi explains that one of the primary frustrations when using artificial intelligence is scattered memory across fragmented tools and isolated chat sessions. When research is conducted across multiple assistants—such as ChatGPT, Claude, and Gemini—none of the platforms retain awareness of what occurred on the others. Furthermore, even within the same platform, context from one specific chat thread does not automatically carry over into the next.
This lack of cross-platform continuity forces users to start their research workflows over repeatedly. Instead of compounding knowledge over time, valuable discoveries become trapped inside abandoned browser tabs and disconnected conversation histories.
Notable Quotes:
"One of the biggest frustrations people have with AI is scattered memory." — Karan Bavandi [0:08]
"AI would not know what went on on another platform." — Karan Bavandi [0:28]
"Moreover, on the same platform, even though there's an overall memory, the context from each chat is not carried over to the next one." — Karan Bavandi [0:36]
[0:59] How Does Kurator Capture and Organize AI Conversations into Structured Markdown?
Speaker: Karan Bavandi
Karan Bavandi demonstrates how Kurator functions as a browser-based web clipper designed to capture research directly from web pages and AI conversations. Clicking the Kurator extension icon opens a popup that automatically scrapes page information in a structured format, enabling users to apply custom AI prompts—such as a research chat ledger—to convert unstructured text into standardized Markdown.
Bavandi explains that the captured content is saved directly into designated folders within the Kurator application, such as dedicated folders for AI conversations or technical discussions. To make knowledge retrieval fast and precise, the system supports publisher tags (identifying sources like ChatGPT, Claude, or Gemini) and content tags that assemble an internal knowledge graph for easy filtering and content repurposing.
Notable Quotes:
"It's accessible on any page in your browser. By simply clicking on the [Kurator] icon, you're going to get a popup that scrapes information off of the page in a structured way" — Karan Bavandi [1:03]
"you can see that it formats content in a structured markdown way customized to how I want to use this knowledge and then save it to a specific folder in my [Kurator] app." — Karan Bavandi [1:32]
"I have tags for publisher so I know these are from [ChatGPT], [Claude], Gemini. I also have content tags so I can very quickly filter information and knowledge graph for my own information." — Karan Bavandi [2:03]
[2:49] How Can You Query Kurator Research Folders in AI Tools Like ChatGPT?
Speaker: Karan Bavandi
Karan Bavandi shows how to connect Kurator research folders directly to AI tools like ChatGPT to analyze collected data and extract actionable insights. In a practical walkthrough, Bavandi prompts ChatGPT to inspect a Kurator folder titled "product research"—which contains curated discussions captured from Reddit and LinkedIn—to summarize why users build second brain systems. ChatGPT scans the entire folder and identifies major user pain points, including fragmented knowledge.
Bavandi also demonstrates querying a "technical discussions" folder containing research on YouTube videos about second brain workflows. By directing ChatGPT to examine these video summaries, he quickly extracts effective hooks and structural breakdowns for producing new video content.
Notable Quotes:
"chat GPT to look at [Kurator] look at product research and [Kurator] and outline the pain points. What are the main reasons people are building a second brain for?" — Karan Bavandi [3:38]
"So it was able to look at my entire research and come up with these pain points. The first one fragmented knowledge which is what I am doing here" — Karan Bavandi [3:48]
"So I asked it look at technical research for people who are talking about their second brain systems and then it gave me the breakdown of what I was looking for." — Karan Bavandi [4:24]
[4:37] Why Is Continuous Context Capture Essential for Building an AI Second Brain?
Speaker: Karan Bavandi
Karan Bavandi concludes that artificial intelligence is far more powerful and productive when users systematically capture their research and supply it as context for future conversations. Feeding pre-structured background knowledge directly into an AI assistant eliminates repetitive setup and ensures prompt responses are grounded in real data.
Bavandi notes that establishing this AI second brain workflow requires no complicated technical setup. Users can install Kurator, define their custom prompts, and immediately begin building a persistent, reusable knowledge base across all their research activities.
Notable Quotes:
"AI is a lot more powerful and productive when you continuously capture your research and use it and as context for your next chat." — Karan Bavandi [4:37]
"what I just showed you, you do not need any setup. You just get [Kurator], create your prompts and start using it to see the real value" — Karan Bavandi [4:59]