Manik_agg

6 ms·

1.
by Manik_agg 5mo ago
Recently started using Codex and Chatgpt again due to claude model getting nerfed or rate limits. Tried gpt5.5 and so far good. Zapier also shared an automation benchmark where 5.5 came on top in the leaderboard https://zapier.co
2.
by Manik_agg 5mo ago
Hey luca, heavy obsidian user here and went through your website and github. Def will try it out. Connecting codex with Tolaria to manage your knowledgebase is something i'm looking forward to try.
3.
by Manik_agg 5mo ago
OpenAI finally catching up with claude
4.
Show HN: Core – open-source AI butler that clears your backlog without you(getcore.me)
3 points by Manik_agg 5mo ago|0 comments
5.
The AI Memory Solution We All Need (No, It's Not OpenClaw)(chrislema.com)
3 points by Manik_agg 8mo ago|3 comments
6.
by Manik_agg 9mo ago
Hey we already had PostgreSQL so no new infrastructure to manage, it was easy way to see if vector database change has any value. It also has good enough performance - handles 10M vectors with HNSW indexes adequately open source - leverage
7.
by Manik_agg 9mo ago
Author here. We've been building CORE (open source) for the past year. Happy to answer questions about the architecture, reification approach, or what broke at scale.
8.
Building AI Memory at 10M+ Nodes: Architecture, Failures, and Lessons(blog.getcore.me)
3 points by Manik_agg 9mo ago|3 comments
9.
by Manik_agg 1y ago
I agree. Asking LLM to write for you is being lazy and it also results in sub-par results (don't know about brain-rot). I also like preparing a draft and using llm for critique, it helps me figure out some blind spots or ways to articu
10.
Show HN: Core – open-source memory graph for AI agents (88.24% SOTA on LoCoMo)(blog.heysol.ai)
4 points by Manik_agg 1y ago|0 comments
11.
How to make Cursor an Agent that Never Forgets and has better project context(redplanethq.ghost.io)
3 points by Manik_agg 1y ago|0 comments
12.
by Manik_agg 1y ago
You’re right dumping all memory into the context window doesn’t scale. But with CORE, we don’t do that. We use a reified knowledge graph for memory, where: Each fact is a first-class node (with timestamp, source, certainty, etc.) - Nodes ar
13.
Ask HN: Can you take your AI's memory with you?
3 points by Manik_agg 1y ago|2 comments
14.
by Manik_agg 1y ago
Claude is incredibly powerful but it's limitation is no persistent memory hence you have to repeat yourself again and again. I integrated Claude with CORE memory MCP, making it an assistant that remembers everything and have a better m
15.
by Manik_agg 1y ago
Figma has come a long way, from a blocked Adobe acquisition to now filing for an IPO.
16.
by Manik_agg 1y ago
Hey - well put! I guess "semantic web" folks were right about the destination, just few years early :P
17.
by Manik_agg 1y ago
Hey - agreed that for basic fact recall, a simple text file + MCP works fine. We designed CORE for complex, evolving memory where text files break down. Example: Health conversations across ChatGPT, Claude, etc. where your parameters change
18.
by Manik_agg 1y ago
Hey we started with llama but since llama was not giving good results hence fall backed to using gpt and launch it. We will evaluate qwen and deepseek going forward, thanks for mentioning.
19.
by Manik_agg 1y ago
Hey - i agree that the demonstrated use can be solved with simple plan.md file in the codebase itself. With use-case we wanted to showcase the shareable aspect of CORE more. The main problem statement we wanted to address was "take you
20.
by Manik_agg 1y ago
Hey plan.md mostly will be a static file that you manually have to maintain. It won't be relational and not be able to form connections between info. You can't recall or query intelligently? (When did my preference change?) CORE l
21.
by Manik_agg 1y ago
Hi, There are 3 major differences between Zep and CORE 1. Market: Zep is B2B focused, CORE indvidual users 2. Portablity: Zep is locked to their platform , CORE works across claude, cursor, windsurf 3. Architecture: Zep is Temporal based vs
22.
by Manik_agg 1y ago
Hey we are actively working on improving support for Llama models. At the moment, CORE does not provide optimal results with Llama-based models, but we are making progress to ensure better compatibility and output in the near future. Also w
23.
Show HN: Core – open source memory graph for LLMs – shareable, user owned(github.com)
112 points by Manik_agg 1y ago|42 comments
24.
Show HN: C.O.R.E – Opensource, user owned, shareable memory for Claude, Cursor(github.com)
20 points by Manik_agg 1y ago|4 comments
25.
by Manik_agg 2y ago
Hey another co-founder of Tegon here, currently we only support Open AI models but plan to add local models support with Olamma soon.
26.
by Manik_agg 2y ago
Hey! Manik here, one of the co-founders of Tegon. Would love to know what are the sources whose context is generally missing in issues in your experience?
27.
by Manik_agg 2y ago
I tried it, looks neat and super helpful for folks who want to improve their english speaking. I know this is early but would have loved an app for the same. Feedback - Can you let me play back the mispronounced word along with what's
28.
by Manik_agg 3y ago
I'm recently started using YNAB via my friends recommendation. The only problem i feel in YNAB is i need to manually do all the entries since I am from India and it only supports US and Europe bank accounts.
29.
by Manik_agg 3y ago
<Data Access/Staleness> Currently every user of the product has to do the auth. However in future for our enterprise customers, we plan to support SSO and SAML. <Google API> You're absolutely right, the array of Google
30.
by Manik_agg 3y ago
Hey I'm one of the co-founders of Poozle. Thanks for asking great questions let me take them one by one <Why LLMs> Our goal is to provide context for LLMs. Our first step is to normalize data and offload syncing, similar to other

More ›