supo

5 ms·

1.
by supo 2mo ago
The environment allows for separation of the useful good ones from the "only in it for the cash" ex-investment banking types ^_^
2.
Against Usefulness(motivenotes.ai)
129 points by supo 2mo ago|40 comments
3.
Show HN: Distributed LLM tracing and GH PR/issue linking [Apache 2.0](github.com)
3 points by supo 3mo ago|0 comments
4.
Like Ollama, but for your own cloud [Apache 2.0](github.com)
4 points by supo 4mo ago|1 comments
5.
by supo 4mo ago
Open-source k8s cluster for serving a wide catalog of small language models across search, OCR & data processing.
6.
by supo 1y ago
by that same logic, why would you not strive to push all the signals you have available into the ANN search? sure, some will have reduced resolution vs using a heavy reranker, but surely the optimal solution is to use the same signals in bo
7.
by supo 1y ago
one thing to remember is that bm25 is purely in the domain of text - the moment any other signal enters in the picture (and it ~always does in sufficiently important systems), bm25 alone can literally have 0 recall.
8.
by supo 1y ago
Here is a quick overview, doesn't really explain the deep details though: https://www.youtube.com/watch?v=ikYsr6nvbdE Basically think mixture of experts, but each expert is an encoder with it's own input tokenizat
9.
by supo 1y ago
This article focuses on ways to make "pre-fetching" more accurate, reducing or eliminating the need for reranking to improve latency/cost but also sometimes quality - for example if you use a text cross-encoder to rerank your
10.
by supo 1y ago
If this was a solved problem, every e-com website would already run some variant of https://arxiv.org/pdf/2209.07663 for all their shopping surfaces and some version of "deep research" on every search query.
11.
by supo 1y ago
It allows faster experimentation because you can't do things like partial embedding updates and reasonable schema migrations on your vector search index - if you could, you'd experiment in retrieval... and with better retrieval yo
12.
by supo 1y ago
If you could wave a magic wand and push all the ranking signals into retrieval and that index would be fast to update and not that expensive to operate - you would do that and you would delete all your reranking systems, wouldn't you?
13.
by supo 2y ago
We are mostly focused on natural language search in e-commerce/marketplace/travel settings right now. There we have a production deployment that drove $15M of incremental revenue through personalization of a shopping feed for a fa
14.
by supo 2y ago
We focus on text, images, numerical, categorical and timestamp-properties for the objects you vectorize with Superlinked. The performance will depend on which models you chose to use with the framework, your queries etc. Happy to elaborate
15.
by supo 2y ago
Supported VDBs: Currently MongoDB, Redis Search and Qdrant https://docs.superlinked.com/run-in-production/index-1 As you say, AstraDB also offers vector search (you can see it at https://superlinked.com&#x2F
16.
Show HN: Superlinked – Vector Embeddings for Structured and Unstructured Data(github.com)
7 points by supo 2y ago|7 comments
17.
by supo 3y ago
A table comparing the features available with different vector databases. The project is open-source and developed by a community of practitioners with points of contact from the different vector databases to validate.
18.
Vector Database Feature Comparison Matrix (35 DBs)(vdbs.superlinked.com)
27 points by supo 3y ago|9 comments
19.
by supo 4y ago
And you decided to not work remotely and 10x your income instantly, because you'd miss your office and colleagues. Got it!
20.
by supo 4y ago
Any apps that can benefit from social interactions. A good example is DuoLingo. They built a social+ product for learning languages, have 50m MAUs and $250M revenue!! There are thousands of other learning apps, but DuoLingo is by far the be
21.
Social features can unlock user engagement and retention(superlinked.com)
6 points by supo 4y ago|3 comments
22.
Next Startup Is You(medium.com)
2 points by supo 5y ago|0 comments
23.
Manage for Impact, Not Performance(svonava.com)
2 points by supo 6y ago|0 comments
24.
by supo 7y ago
The removal of barriers favors those who can take advantage of the newly acquired access and information. Your life reflects this capacity as much as the environment that let it manifest itself:)
25.
Googler Quits to Join a Startup Fixing VC(svonava.com)
1 points by supo 7y ago|0 comments
26.
Googler Advice to a Fintech Startup Founder(svonava.com)
3 points by supo 7y ago|0 comments
27.
Goodbye Google and Salve Startup(svonava.com)
2 points by supo 8y ago|0 comments
28.
Cryptocurrency correlations over time (building a decorrelated portfolio)(svonava.com)
3 points by supo 9y ago|0 comments
29.
by supo 9y ago
Nice formalization, this has been rattling around in my head for a long time it's good to see numbers put on it!
30.
by supo 9y ago
A short piece of satire on the dangers of deploying ML/AI into production.

More ›