Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
cheesyFishes
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
1.
▲
by
cheesyFishes
1y ago
Thanks! Seems like CLI tools are all the rage these days
2.
▲
Workflows 1.0: A Lightweight Framework for Agentic systems
(llamaindex.ai)
10 points
by
cheesyFishes
1y ago
|
2 comments
3.
▲
Bending without Breaking: optimal design patterns for effective agents
(llamaindex.ai)
6 points
by
cheesyFishes
1y ago
|
0 comments
4.
▲
by
cheesyFishes
2y ago
Pretty neat initial launch. What's top of mind to add to it?
5.
▲
by
cheesyFishes
3y ago
There is a PDF parser, LlamaParse, (which is open to everyone), and a managed ingestion/retrieval service, that is currently invite-only. Planning broader releases in the future for sure.
6.
▲
by
cheesyFishes
3y ago
The output is either text or markdown, and from there you can handle it however you need. In LlamaIndex for example, there are a a few markdown-specific classes that work well with this. You can find an example over in the repo -- https:&#
7.
▲
by
cheesyFishes
3y ago
This is actually what llama-index does if you don't use a vectordb integration
8.
▲
by
cheesyFishes
3y ago
The real advantage of LLMs is not search -- that's just something that's easy to show people. No one really wants to talk to an LLM instead of using Google search. The real value is being general reasoning machines. Being able to