Memori v3 Private Beta β Call for Early Testers π #176
Replies: 15 comments 17 replies
-
|
happy to see more memory solutions appear which will led to simplifying context engineering username: yusuf-eren |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
Tirth-1999 |
Beta Was this translation helpful? Give feedback.
-
@Tirth-1999 Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
Love the mission and would like to contribute |
Beta Was this translation helpful? Give feedback.
-
@Vaishvi12 Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
Awesome project! username: triadcode |
Beta Was this translation helpful? Give feedback.
-
@triadcode Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
padamshrestha |
Beta Was this translation helpful? Give feedback.
-
@padamshrestha Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
PureTech |
Beta Was this translation helpful? Give feedback.
-
@danielpennick Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
briangladu |
Beta Was this translation helpful? Give feedback.
-
@briangladu Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
Great π |
Beta Was this translation helpful? Give feedback.
-
@akenterprises47-dev Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
KumarBalajiNagaraj |
Beta Was this translation helpful? Give feedback.
-
@KumarBalajiNagaraj Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
Wanna configure it with Gemini... |
Beta Was this translation helpful? Give feedback.
-
@Priyansxu Thanks for your interest! We published the beta release package on PyPI: https://pypi.org/project/memori/ Feel free to get and install it! I would love to get your feedback. |
Beta Was this translation helpful? Give feedback.
-
|
Yes, it should work with Gemini |
Beta Was this translation helpful? Give feedback.
-
|
Your heart is very big. Thank you. |
Beta Was this translation helpful? Give feedback.
-
|
excellent work bob! happy to test it out and will let you know the results soon on our latest release - memorisdk 2.3.3 i will be using mongodb database with these features enabled for my agentic RAG based chat application : MEMORI_CONSCIOUS_INGEST=trueMEMORI_AUTO_INGEST=true(long term memory,short term memory and fact extraction) embedding model : OpenAI |
Beta Was this translation helpful? Give feedback.
-
|
The key thing that makes me very interested in this project is optional vector database retrieval - saving lot of cost and latency for user responses of full embedding, retrieval and re-ranker pipeline along with saved llm call with all knowledge base chunks |
Beta Was this translation helpful? Give feedback.
-
@Gentoflakes Thanks a lot! Thatβs exactly what Memori is built for. Removing the vector DB pipeline saves huge cost and latency:) Curious to know: what use case or project are you working on where this would help the most? |
Beta Was this translation helpful? Give feedback.
-
|
I am building multiple solutions which include ai agents, prompt library, agentic RAG chat best use cases
among many others but above are the ones im starting with. Will be happy to hear your thoughts and take on this |
Beta Was this translation helpful? Give feedback.
-
|
HI @Boburmirzo, I'm building an AI voice agent that works as a preparation simulator for a candidate who will get examined in the MRCPCH clinical exam. I've made the AI part successfully. The tough part is to evaluate the candidate against a rubric. I don't want to build rule engines or something like this, and I think Memori will help to achieve this. I need something that has the context of LLM response and user response, but evaluate it against a rubric. Does Memori help in that? |
Beta Was this translation helpful? Give feedback.
-
@MahmoudHousam Thanks for the question! And yes, Memori v3 can help with this, depending on how you want to structure the evaluation. I might be wrong, @mcmontero can correct me if I am wrong, but memori doesnβt directly βscoreβ a user, but it gives your agent the structured memory needed to score. For example:
This gives the LLM a clean, structured representation of what the candidate did, instead of relying on raw conversation transcripts. |
Beta Was this translation helpful? Give feedback.
-
|
@Boburmirzo Thanks for your support. I don't need score eventually. The exact example you mention is what I'm looking for. So, Memori can automatically extracts the required info I pass to it or it has some kind of schema to extract relevant info? If you have a working example, this would be great |
Beta Was this translation helpful? Give feedback.
-
@MahmoudHousam Yes, it should work out of the box. I tried to implement something small here: https://github.com/MemoriLabs/python-sdk/tree/example-clinical-exam/examples/clinical-exam-evaluator Could you run this example and see if it fits your needs? |
Beta Was this translation helpful? Give feedback.
-
|
memory really needed for ai enhancement in the future username: diwanmuhamad |
Beta Was this translation helpful? Give feedback.
-
|
ΰΈΰΈ‘ΰΈΰΈ³ΰΈ₯ΰΈ±ΰΈΰΈΰΈ±ΰΈΰΈΰΈ²ΰΈ£androidΰΈΰΈ’ΰΈΉΰΉΰΈΰΈ±ΰΈΰΈΰΈ³ΰΈΰΈΰΉΰΈΰΈ΅ΰΈ’ΰΈ§ΰΉΰΈ‘ΰΉΰΈΰΈ±ΰΈΰΈͺΰΉΰΈΰΈΰΈ²ΰΈ iPhone ΰΈΰΈΰΈΰΈΰΈ±ΰΈΰΉΰΈ‘ΰΈ·ΰΉΰΈ 26 ΰΈ.ΰΈ’. 2568 ΰΉΰΈ§ΰΈ₯ΰΈ² 19:39 ΰΉΰΈΰΈ΅ΰΈ’ΰΈΰΉΰΈΰΈ’ Bobur Umurzokov ***@***.***>:ο»Ώ
HI @Boburmirzo, I'm building an AI voice agent that works as a preparation simulator for a candidate who will get examined in the MRCPCH clinical exam. I've made the AI part successfully. The tough part is to evaluate the candidate against a rubric. I don't want to build rule engines or something like this, and I think Memori will help to achieve this. I need something that has the context of LLM response and user response, but evaluate it against a rubric. Does Memori help in that?
@MahmoudHousam Thanks for the question! And yes, Memori v3 can help with this, depending on how you want to structure the evaluation.
I might be wrong, @mcmontero can correct me if I am wrong, but memori doesnβt directly βscoreβ a user, but it gives your agent the structured memory needed to score.
For example:
Candidate answers a clinical question
Memori automatically extracts skills, facts, events from that answer
Your next LLM call can retrieve only the relevant memories
The LLM can then evaluate the candidate using the rubric + the retrieved structured memory
This gives the LLM a clean, structured representation of what the candidate did, instead of relying on raw conversation transcripts.
βReply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Hey everyone! π
We are excited to announce the private beta for Memori v3, our biggest upgrade yet to the memory fabric for enterprise AI.
Memori v3 is designed to fit directly into the software and infrastructure you already use: LLM-agnostic, datastore-agnostic, and framework-agnostic, with seamless integration into any architecture.
π₯ Whatβs New in Memori 3?
Weβre inviting a small group of early testers to help us validate and refine the new release before it goes public.
We published the beta release package on PyPI:
https://pypi.org/project/memori/
Feel free to get and install the latest version! I would love to get your feedback. Please share your feedback in the comments below!
Thanks for helping us shape the future of Memori.
This is a major step forward, and weβd love to build it together with you. β€οΈ
Beta Was this translation helpful? Give feedback.
All reactions