Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
krawczstef
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
krawczstef
3mo ago
FYI - Burr is designed with recursion in mind, i.e. the ability to kick off Burr within Burr. So you could have an action that is managing several Burr subgraphs... Or you write your own management layer here.
2.
▲
by
krawczstef
3mo ago
yep - here's something cool an end user wrote with Burr - https://github.com/msradam/phoebe Burr just helps you, the engineer, to really control the primitives. Then adds some cool features you don't have to
3.
▲
by
krawczstef
3mo ago
no. more lower level.
4.
▲
by
krawczstef
3mo ago
user contributed :)
5.
▲
by
krawczstef
3mo ago
Cause I submitted it. Learning the Apache process and cranking on other things has been a slow process. But we've got some momentum and beginning more regular releases.
6.
▲
by
krawczstef
1y ago
I agree with this post. Otherwise the larger picture is that MCP is a land grab for building an eco-system around integrations to get access to data. Your LLM agent is not valuable if it can't access things for you... and from a market
7.
▲
by
krawczstef
2y ago
Unfortunately their blog doesn't have the example they have on x.com - https://x.com/ZyphraAI/status/1888996367923888341 - which I think is more illustrative.
8.
▲
Test Driven Development of LLM / Agent Applications with Pytest
(blog.dagworks.io)
2 points
by
krawczstef
2y ago
|
0 comments
9.
▲
by
krawczstef
2y ago
your comment made me think I read it wrong, but it's not closed source, they're just moving to a source available license...
10.
▲
by
krawczstef
2y ago
yep, agreed. I think that's another way to view the current state of "GenAI" tooling (e.g. all those complicated frameworks that received $M in funding) and why things like https://www.anthropic.com/research&#
11.
▲
by
krawczstef
2y ago
+1 for vanilla code without LangChain.
12.
▲
by
krawczstef
2y ago
Interesting! I feel like this is a cross between https://github.com/dagworks-inc/burr (switch state for context) and https://github.com/Netflix/metaflow because the output of the "task"
13.
▲
by
krawczstef
2y ago
This is reminiscent of themes brought up in https://news.ycombinator.com/item?id=42253735 and most famously in https://news.ycombinator.com/item?id=36645575 ...
14.
▲
The graduation problem: avoid frameworks getting in the way
(blog.dagworks.io)
5 points
by
krawczstef
2y ago
|
1 comments
15.
▲
by
krawczstef
2y ago
Forgot to mention, on the roadmap we have pluggable executors, e.g. delegating parallelism to whole agents/subgraphs to systems like "ray". Would love to collaborate with folks if that's interesting.
16.
▲
Show HN: Modeling parallel (and hierarchical) agents with Burr
(colab.research.google.com)
2 points
by
krawczstef
2y ago
|
1 comments
17.
▲
by
krawczstef
2y ago
Looks great. But being burned by other "abstractions", e.g. LangChain, I'm weary of the oversimplification. How are you not going to make those same mistakes?
18.
▲
by
krawczstef
2y ago
Does this compete with Minio?
19.
▲
by
krawczstef
2y ago
@staranjeet we should get you plugged in as part of persistence with https://github.com/dagworks-inc/burr !
20.
▲
by
krawczstef
2y ago
Yep. When I was designing https://github.com/dagworks-inc/hamilton part of the idea was to make it easy to understand what and where. That is, enable one to grep for function definitions and their downstream use easily
21.
▲
by
krawczstef
2y ago
Wow. Implemented pregel to run a 3 node graph... Jokes aside, that's an interesting thesis to have in there...
22.
▲
by
krawczstef
2y ago
Really? Can you point to code for that? To me it's just a way to describe a graph and execute it, exactly like https://github.com/dagworks-inc/burr (I'm an author).
23.
▲
Show HN: Streaming Chatbot with Burr, FastAPI, and React
(blog.dagworks.io)
1 points
by
krawczstef
2y ago
|
0 comments
24.
▲
by
krawczstef
2y ago
Here's a blog post that I just pushed that doesn't use them at all - https://blog.dagworks.io/p/building-a-conversational-graphdb (we have more on our blog - search for RAG). [disclaimer I created Hamilton &a
25.
▲
by
krawczstef
2y ago
getting into enterprise is hard, so probably trying open source to help with that.
26.
▲
by
krawczstef
2y ago
What would you like to do, but can't? BSD-3 Clear Clause is very permissive - which covers the majority of the code. What isn't BSD-3 is around features that are targeted at enterprises, e.g. auth. But you don't need to use t
27.
▲
by
krawczstef
2y ago
Awesome! Great to hear! Would love to get your thoughts on the UI :) Feel free to file issues.
28.
▲
by
krawczstef
2y ago
thank you!
29.
▲
by
krawczstef
2y ago
Thanks!
30.
▲
by
krawczstef
2y ago
This is what I use google keep ( https://keep.google.com/ ) for -- though it's not as fully featured as yours, but it is free. In your opinion, what would be the main feature over that you'd say?
More ›