ankrgyl

7 ms·

31.
by ankrgyl 4y ago
Thank you ultrasounder! Supply chain, construction, purchasing, insurance, financial services, and healthcare are our biggest verticals. Although we have customers doing just about anything you can imagine with documents! For invoices, we h
32.
by ankrgyl 4y ago
It's very easy to try! Just plug that URL here: https://huggingface.co/spaces/impira/docquery . I tried a few questions: What is the development date? -> June 20, 2017 What is the medicine? -> SPIN
33.
by ankrgyl 4y ago
> but my use case is in creating an offline-first local document storage system (data never reaches a cloud). Makes sense -- this is why we OSS'd DocQuery :) > The templates sound very cool - are they essentially just using a pre
34.
by ankrgyl 4y ago
Can you tell me a bit more about your use case? A few things that come to mind: - There are some ML/transformer-based methods for extracting a known schema (e.g. NER) or an unknown schema (e.g. relation extraction). - We're going
35.
by ankrgyl 4y ago
DocQuery ( https://github.com/impira/docquery ), a project I work on, allows you to do something similar, but search over semantic information in the PDF files (using a large language model that is pre-trained to query
36.
Show HN: DocQuery, an OSS tool for analyzing documents with LLMs(huggingface.co)
7 points by ankrgyl 4y ago|0 comments
37.
by ankrgyl 4y ago
Many of the comments seem to miss the point entirely -- this post is about how to achieve sufficient decentralization: > two users can find each other and communicate, even if the rest of the network wants to prevent it and use central
38.
by ankrgyl 4y ago
The more languages you learn, especially with reusable concepts, the less of a tax it is to learn a new one. As someone who has written C/C++ for over a decade, learning Rust felt very straightforward -- I just had to learn about the b
39.
by ankrgyl 4y ago
It allows you to create many more language threads (go routines) than kernel threads, and unlike other languages properly hides this abstraction from you with a rope stack (so you don't need to create coroutines or use async/await
40.
by ankrgyl 4y ago
It's not that there's an "advantage" to pointers, but rather it's the reality that they exist. Everything you're doing in JS also "uses" pointers under the hood, but the language and runtime abstract
41.
by ankrgyl 4y ago
Pointers :). C++ and Rust try to hide the concept away with References, which add a very helpful amount of type safety, but do not protect you from the fundamental idea of allocating memory and then later free-ing it. It's not that you
42.
by ankrgyl 4y ago
In that case, I'd suggest learning Go first, as a kind of "step up" to learning Rust. If you haven't already, I'd also suggest learning C. I think there's some linearity to these tasks. Rust (and C++) are an am
43.
by ankrgyl 4y ago
If your goal is to be a better coder long term, I’d suggest learning both. As many have pointed out, Go is quick to learn, and I think learning both will only take a small amount of extra time (compared to learning Rust). They each present
44.
by ankrgyl 4y ago
If you want to play with MDX without having to set it up, check out https://motif.land . It comes with a collaborative editing environment too, that is incredibly easy to use for authoring markdown (our non-technical folks use it
45.
by ankrgyl 4y ago
This is mind blowing stuff. I was skeptical that you could use the browser effectively as an intermediary to the file system, and that CRDTs would work reasonably well on files, but they seem to have overcome both obstacles. I’m curious, ho
46.
by ankrgyl 4y ago
At the very least, that does not work for rolling back transactions with the database/sql ( https://pkg.go.dev/database/sql ) package, although it may work for other cases. We've had numerous production bugs re
47.
by ankrgyl 5y ago
We let the candidate do whatever they'd like during the interview. Some liked to talk out loud, and others preferred not to. We didn't care. Also — we gave candidates one hour (the author of the blog post does not seem to remember
48.
by ankrgyl 5y ago
I had never looked at memcached's source code prior to thinking of the question, but we did try it out before (and measure the time) before administering it for the first time.
49.
by ankrgyl 5y ago
Luckily I did not have to deal with that during my tenure (I left in 2016). However, we had experimented with other changes to memcached as well as similar changes to other codebases. There is no shortage of OSS projects to draw from. Howev
50.
by ankrgyl 5y ago
It was a lot harder (even at MemSQL) to repeat the success of this question. At Impira, it took us about a month of trial and error over 5-6 similar codebases to find a good (and appropriately relevant) question.
51.
by ankrgyl 5y ago
Yes it did, at first. It was one of the first times we presented the question, so we didn't have a great signal on its efficacy just yet. However, other red flags emerged during the interview process, and we decided to try the question
52.
by ankrgyl 5y ago
We did hire people that failed the test.
53.
by ankrgyl 5y ago
Yes definitely -- although it's pretty crazy how much not being under pressure affects your ability to complete the question in X minutes. I've learned over time that good interviews take a lot of experimentation, and the best t
54.
by ankrgyl 5y ago
It's equally valid — your reasoning exceeds our expectation for the question. Essentially we're trying to test for whether you can understand at least one correct implementation. Whether or not your approach is "better"
55.
by ankrgyl 5y ago
You can't take that for granted, but it turns out in practice that memcached does not use atomic add, and so it happens to be straightforward to support multiplication with the same set of locks.
56.
by ankrgyl 5y ago
I got super lucky and came up with the question an hour before I first administered the interview. Memcached was the first codebase I went for, because someone had asked me a question involving it a few months ago, and it was conceptually a
57.
by ankrgyl 5y ago
The rubric was more of a "Y/N" for this question. But we'd refer to the author's solution as "great". Messier solutions would involve a lot of copy/pasta and occasionally inefficient or incorrect impl
58.
by ankrgyl 5y ago
At the time we debuted the question, it was actually a recruiting advantage for us. People would interview with us and other "big tech" firms, and get excited about the opportunity because we asked this question.
59.
by ankrgyl 5y ago
Fun fact: we tried it out with Redis, but the question worked out to be too easy because of how nicely the codebase is laid out.
60.
by ankrgyl 5y ago
I think "necessary but not sufficient" is an accurate way of capturing our goals for the interview question. Some really great candidates felt just like you do about the question, and we learned over time how to present it to seni

More ›