gdad

4 ms·

1.
by gdad 1mo ago
Haha! I am going to put this one up as a win! Thanks for reading! Hope you found it useful.
2.
by gdad 1mo ago
Truly. Doing this for coding agents is an interesting and different shaped problem.
3.
by gdad 1mo ago
Thanks Samyakk! 1. Yes, works on docs, agent conversations, human-conversations from different sources (Slack, JIRA, etc.). We have connectors for some of these as well; so it is plug and play 2. conventional RAG recall accuracy is quite lo
4.
by gdad 1mo ago
Interesting. Where can I read more about this?
5.
by gdad 1mo ago
Apologies. Should be working now.
6.
by gdad 1mo ago
I also wrote a shorter preview here: https://www.maximem.ai/blog/agentic-context-management-paper
7.
Agentic Context Management: Memory and Cost as Architecture Problems(arxiv.org)
79 points by gdad 1mo ago|28 comments
8.
Treating Agent Memory and Costs as Lifecycle and Architecture Problems(arxiv.org)
2 points by gdad 2mo ago|0 comments
9.
by gdad 4mo ago
Aiden, Claude Code
10.
by gdad 7mo ago
Building in the same space (Maximem Vity, cross-LLM memory as a Chrome extension and OpenClaw plugin). The SMTP analogy really resonates with me. I wrote a comparison a few weeks ago of how ChatGPT, Claude, and OpenClaw actually implement m
11.
by gdad 8mo ago
Author here. I have been learning about how ChatGPT, Claude, and OpenClaw handle persistent memory; what actually gets stored, how it is retrieved, etc. Quick summary of the findings: - ChatGPT pre-computes lightweight summaries of your ~15
12.
Comparing how 3 AI assistants implement memory(maximem.ai)
1 points by gdad 8mo ago|1 comments
13.
by gdad 8mo ago
really helpful pov.
14.
by gdad 8mo ago
doesn't that put people off? Like in early conversations?
15.
by gdad 8mo ago
aligned. And that is also what i am focussed on. But a friend suggested this as an optionality and I am wondering if it makes sense.
16.
by gdad 8mo ago
Why do you say so? Software patents are rarely guarded? Or something else?
17.
by gdad 8mo ago
Makes sense (do believe something big is on the cards). As a micro-entity, filing fees are nominal, so I guess it doesn't hurt.
18.
Ask HN: Do provisional patents matter for early-stage startups?
20 points by gdad 8mo ago|20 comments
19.
by gdad 8mo ago
I experienced this as well. The built-in memory features are a black-box. Can't see what's stored, can't control when it's recalled and takes random irrelevant examples from past conversations into unrelated conversation
20.
by gdad 8mo ago
I tried it last Monday and was in awe. All my team members now use it. I for one, am stepping cautiously around it. And hence had similar questions. Yesterday, we released a memory plugin for it (based on their plugins framework). And we go
21.
Show HN: Memory plugin for OpenClaw; cross-platform context sync with major LLMs(memoryplugin-for-openclaw.com)
2 points by gdad 8mo ago|0 comments
22.
Ran a 5k queries on 50k documents to understand the file vs. vector RAG debate
2 points by gdad 9mo ago|0 comments