jonnycoder

I just learned React/Nextjs for a take home coding interview assignment. I used useState a few times. Why would I need redux when there’s useState and useContext?

Other than being overly verbose, I think React is ok after building my website with Sveltekit. There just seems to be way more support for off the shelf react components and 99% of jobs ask for React.

I wrote most of the VIZIO code for consenting and turning on/off the Viewing Data, Terms and Conditions and Privacy Policy. You can easily toggle that stuff in the settings menu and during the Out-of-box-experience.

Last I was there was over a year ago, but VIZIO was one of the first to get in big trouble with the US government due to privacy issues back around 2016 I think? It's on wikipedia if you are curious.

Hey I worked on that project along with bluetooth voice remote. I personally spent a few weeks during Christmas time (2020 or 2019?) making sure the bluetooth pairing process was rock solid. Ending up root causing two deadlock bugs that had a repro rate of like 1 in 50.

The engineering leadership was going downhill with layoffs and politics. We would have the quality of the firmware/smartcast be rock solid and then following quarter it took a nose dive. Some of the best people I ever worked with are no longer there.

And last I know, VIZIO isn't like samsung and tries to randomly find an internet connection to update firmware. I still use my free 70in with Apple TV and it is great because I disconnected it from the internet (after I was laid off).

I wrote a review about Replicate last week and cog I was using, insanely-fast-whisper, had boot times exceeding 4 minutes. I wish there was more we can observe to find out the cause of the slow start up times. I was suspecting it was dependencies.

https://open.substack.com/pub/jonolson/p/replicatecom-review...

On a related note, I recently learned about the got-scraping module which doesn’t use chromium or any browser but its good at mimicking a browser and executes javascript. I also wrote a module that parallelizes browserless.io / playwright and makes it really cheap to use a cloud scraping solution.

I have the same experience. I've been unemployed for 5 months as a Senior Software Engineer and I hit most of the job requirements. The only skill I don't directly match is React development for full stack jobs, but I am building an LLM app in React so that I can fix that hole. Speaking of LLM, I'm surprised I'm not seeing any job skill requirements for LLM related topics or even vector database experience. Are the bigger companies so far behind the curve here?

This would have been great when I used to work in embedded development and had to grep log files to root cause bugs. I had thought about writing something like this but was always too busy. The biggest feature is combining log files and organizing by timestamp.. well done

I'm looking for remote dev job, which is much harder, since I am in Sacramento.

That's a good point about the applicant number. I don't think anyone knows exactly how it works, but it was the first time I saw a job posting older than a few hours with <10 applicants with such straightforward skills such as Python.

"So It works well for someone smart who is doing their own research."

That's a concise explanation that also applies to GPTs and software engineering. GPT4 boosts my productivity as a software engineer because it helps me travel the path quicker. Most generated code snippets need a lot of work because I'm prodding it for specific use cases and it fails. It's perfect as an assistant though.

Thanks. I'm starting to realize that part of the problem is job search and matching.

I was contacted by a company recruiter for a small healthcare SaaS in California and had 3 interviews recently. When I looked up the job, only 7 people had applied in 2 weeks on LinkedIn. They are a very real company with very real people, but their job post is not getting seen (it's not a promoted post).

My next AI project will be to scrape LinkedIn jobs, analyze it for repost/promoted behavior, group it by consulting/headhunters vs company job post, eliminate duplicates, and filter based on my skillset and hard-no qualities (such as can't work if I live in California, must be in EST but I'm in PST timezone, requires Java experience, etc).

Thanks. This is the first real test apart from a couple dozen test users. I've received hundreds of prompts in the past 24 hours from Hacker News users, mainly from my suggested questions buttons.

The actual questions I got did not provide a response that is to my liking. Most of that is due in part because I'm using gpt3.5 since gpt4-turbo is a lot more expensive, and I can learn a lot more by using an inferior LLM.

For example, using an llm router to analyze the query and route to a specific helper function with a specific prompt would be helpful. Sometimes a user starts with a greeting but the response is a pre-written "Sorry an answer cannot be found". Questions are typically grouped into a category such as skills, experience, project, personal (ie: where are you located), preferences (ie: favorite language), and general interview questions (ie: why should I hire you). Questions in categories can be better answered by using a different prompt and/or RAG technique.

This is very easy to do and a great idea!

Langchain and Llama Index both have classes to read a directory, if you git clone, and perform a RAG.

If you want it to scrape a github URL, there is a module for that too!

This starter tutorial will do RAG on a directory of files: https://docs.llamaindex.ai/en/latest/getting_started/starter...

Very cool, I signed up. I agree that practicing a coding interview is better under pressure. It's a much difference skill to solve a coding problem both under time pressure and pressure to speak your thoughts to entertain the interviewer. Only practice can help improve that skill.

I built an Interactive Resume AI chatbot where anyone can ask questions about my experience and skills: https://www.jon-olson.com/resume_ai/

The backend is a Python FastAPI that uses ChromaDB to store my resume and Q&A pairs, OpenAI, and Airtable to log requests and responses. The UI is Sveltekit.

I'm currently building a different tool and will apply some learnings to my Interactive Resume AI. Instead of Airtable, I am going to use LangSmith for observability.

I started writing and my Substack articles are also linked to via my website. I'm currently working on applying sentence window retrieval and that article will be out shortly. This is part of a #buildinpublic effort to help build my brand as well.

I've been unemployed since Sept as a Senior Software Engineer. The market is tough so I'm focusing on the above to help get employment or a contract.

Also heard about shipfast on twitter and reddit. Twitter is big for promoting solopreneur stuff like this.

On a side note, LinkedIn has a lot of frequent posts if you follow certain projects. LlamaIndex is killing it on this front. Lots of great stuff and they even teamed up with Andrew Ng with advanced RAG techniques on deeplearning.ai

I was just starting to build my interactive resume ai when I saw anyeung's original post about his product here: https://news.ycombinator.com/item?id=38245665

I signed up for his service to learn more about it. We actually had a few technical calls. He is a really smart guy, very focused.

I did eventually finish my interactive resume ai found at https://www.jon-olson.com/resume_ai

His is still better but I'm improving/upgrading my code to use LlamaIndex with some advanced RAG techniques.

Yes, there is a lot written about it. Here is one link I have saved:

https://github.com/niespodd/browser-fingerprinting

Man I sure hope so. I spent the last month or two working on my portfolio website to help stand out. I built an interactive resume ai, mostly as novelty, but to help get back into frontend dev. I'm building projects, writing tech articles and engaging on twitter with fellow solopreneurs/entrepreneurs. My worst case scenario plan is to build something that makes me money eventually.

Leetcode is not superficial when nearly all companies are using leetcode as a barrier to entry these days. Many of us with 10+ years of experience did not need to practice leetcode to get hired originally, but many would fail miserably today due to the time constraint and necessity to socialize/speak the problem while solving it in 45 min. Some practice problems took me all day at first, but after a while it became a skill of quickly identifying the type of data structure to use and solving it. It's great practice.

As for LLMs, it's probably the best thing to learn right now, especially by building something. I personally believe it will lead to an explosion in potential jobs. It's hard to describe but imagine a software engineer learning how to create a rest api for the first time after doing soap. And then integrating various apis to harness saas products, cloud deployments, etc back in the 2000s.

Then again I also talk to an ex-googler I met on Hacker News who built an LLM app and is now spending most of his time selling his saas product because he can't find a job either. That's a very positive go-getter attitude.

Leetcode is good practice but go into full exploring mode and learn new tech. I got back into full stack development (mainly backend guy) and I now have endless ideas. I learned sveltekit and tailwind css to build my portfolio website. I love rucking and did some research and found out there's no real good rucking tracker. Consider social media such as Twitter and Reddit groups like saas, entrepreneur, and indiehackers.

This article blew my mind because we as software engineers have immense power, we just need to use our skills to build stuff and take advantage of social media to harness/sell it. https://www.henrikkarlsson.xyz/p/search-query

Same, 15 years experience and the only major programming languages I don't have experience in are Java and Go.

The job postings are there, but they sure don't seem real. I've had 1 interview in past 3 months and that's only due to being internally referred at NVIDIA for about 20+ matching jobs.

I am a Senior/Staff level software engineer who has been unemployed for 4 months. Best advice is to live way below your means and have great savings. I get a little freaked out at the job market and not having all but 1 interview in the past few months, but I don't feel that stressed because I'm ok financially for now.

I'm an unemployed staff/sr level software engineer who's been unemployed for 4 months. The only interview I got in the past 3 months was due to an internal referral. I'm seeing a lot of similar re-posts.

I’ve applied to 20+ nvidia sr software engineer jobs since february of last year and every single one had an internal manager referral (former colleague at another company).

I got 3 interviews. The hiring managers were nice, but they just had so many other applicants to choose from that I never made it past the first interview.

This article really resonates with me. I’ve spent the last two months creating a personal/portfolio website, developing an interactive resume ai, creating an open source web scraping utility, and researching tech non stop.

I would say I’ve been 80% exploring and 20% exploiting/doing. Everyday I come across tech that could help me build something even more efficiently. Examples include plasmic.app, directus.io, and tools for SEO combined with ideas of automation.

But I also am left with the feeling that I need to switch gears and do 80% doing and focus on just that one project. It’s just hard finding an ideal project/saas idea to commit to because it’s a common theme that we fail much more than we succeed. But as the author pointed out, that’s just another person’s mental model. I really need to find out myself.

“Retooling, that’s a good excuse.

I’LL RETOOL YOU!”

“Dear loyal customers, in these robust economic times we regret to inform you that we are laying you off. You have 3 months of support severance.”

I posted the short courses just as answer to how to address quality control. I'm not selling anything, and those courses are free anyway. deeplearning.ai was cofounded by Andrew Ng, who is probably the most well known for his work on teaching machine learning through deeplearning.ai, Coursera, Stanford, etc. He has taught and influenced millions.

https://en.wikipedia.org/wiki/Andrew_Ng

In regards to "evaluation", I think these is what those short courses will cover:

Self-Evaluation with the LLM: The idea is to use the language model to generate an answer and then use the same or a different model to evaluate that answer. The evaluation could involve asking the model to rate the answer's accuracy, coherence, relevance, or any other desired metric. This self-evaluation process can be automated and scaled, although it's important to be aware of the limitations, as the model might inherit biases or blind spots from its training data.

LangChain for Structured Evaluation: LangChain can be used to structure this self-evaluation process. It can orchestrate the flow where the LLM first generates an answer and then follows a series of steps to evaluate it. This might include breaking down the evaluation into specific questions or tasks that the LLM must perform to assess its initial response.

The quality control is a good question, and one that can probably be addressed using evaluation as taught by some of the deeplearning.ai short courses (1).

I made an interactive resume ai bot on my personal website and there is an instance where I can ask it "tell me about your intel experience" and it added in C++ as one of the languages, but that is untrue. I had done C++ at a different company.

1. https://www.deeplearning.ai/short-courses/