samjbobb

Kaavio | https://kaavio.ai | REMOTE (US or Canada) | Full Time

Kaavio builds AI-powered tools that help people buy and sell complex physical products. We work with industrial and science distributors managing massive catalogs (millions of SKUs, each with detailed specs and docs), extracting images, descriptions, and structured data for 10k+ products in days instead of months.

Hiring engineering positions 2 and 3:

Applied AI Engineer: work on research agents, RAG over large catalogs, structured extraction, and the eval work that makes AI trustworthy in production.

Data Engineer / Backend Engineer: work on the data platform, durable workflows, APIs, and infrastructure that keep a data-intensive AI app fast, reliable, and affordable.

$180k–$220k base, 0.25%–0.75% equity, depending on experience

More role info and application instructions here: https://github.com/KaavioAI/jobs

This was a great read. Thanks to the author for the vulnerability.

In this thread and others, HN does a lot of talking about the fixed properties of these languages and frameworks.

Something I don’t see much is: choose the tool that you and your team are most familiar with.

It’s relatively easy to learn a new language, but it takes a long time to learn an ecosystem. Learning that ecosystem takes time away from focusing on your customer.

I recently started a new web app and tried to use Django because of the batteries-included framework “should” have been the right tool for the job.

But I haven’t used Django in 15 years and it was slow going. After a few weeks of slow going, I rewrote the whole thing in Typescript in a week. Not because of any general argument that TS is better, but because it’s the ecosystem I know well. Not by choice, but as a function of the jobs I’ve had.

I’d like to suggest that it’s okay to choose the tool you’re most familiar with, because you’ll move faster, write better code, and frankly web apps can be written in any of these languages.

Oh wow. I love the idea of DevLog. I’ve spent a lot of time helping engineering teams get better at regularly communicating what we’re doing, with context, that everyone in the company can understand. Have you shared this yet?

The indication for EoE is weekly, though I seem to be able to get away with every other week.

Dupixent is a subcutaneous auto-injector pen, so it’s the least-bad injection.

Injections definitely suck and everybody has to make their own trade-off assessment.

If you can’t figure out your triggers, and a lot of people can’t, it’s great that we have an effective therapy available now. It dramatically improved my life … once I got over injection anxiety

FYI for any other EoE folks: Dupixent is now indicated for EoE. It’s resolved my symptoms completely (both the ones I notice and the ones my gastroenterologist checks via scope). I don’t experience any side effects.

It’s an off-label use, but swallowed fluticasone is also effective for many people. Also has no to minimal side effects. I did that for many years before dupixent was available.

Exciting that you figured out your trigger. I collected a lot of data and made a lot of graphs, but never found anything conclusive.

Great article. Taking control of your health via the tools we have as technical professionals is awesome.

Here’s a guide on exploring 120 years of global weather measurements with Clickhouse. I enjoyed following this guide over a weekend. Also it gives you a sense of the dataset, even if you’re not interested in the clickhouse part. https://clickhouse.com/docs/en/getting-started/example-datas...

Copia Automation | Sr Software Engineer, Head of Product | Remote + NYC | copia.io

Want to work on building DevOps tools for industrial automation? We’re building better code and ops tools for factories, warehouses, and anywhere that the physical world is automated.

Check out https://www.copia.io/careers

Copia Automation | Sr Software Engineer, Staff Software Engineer, Head of Product | Remote + NYC | copia.io

Want to work on building DevOps tools for industrial automation? We’re building better code and ops tools for factories, warehouses, and anywhere that the physical world is automated.

Check out https://www.copia.io/careers

Kiwix is what you want.

I keep offline copies of Wikipedia and the major StackOverflow sites.

https://www.kiwix.org/en/

Impressive results. Nice write up

This is very interesting. I’ve been building an internal system that looks a lot like this for the current startup I’m at.

Will be following.

I’ve been thinking about exactly this!

I really miss the ‘low effort’ social interactions that in-office work provides. “Want to get lunch?” is one of those.

Question: does this app already exist? Can I facilitate this in my life without building the app (easy part) and building a significant user base (hard part)?

I think this is reasonable advice, in some settings. But for many of us, I think it’s just not practical anymore.

The lines have become too blurred. I work from home, I have one office and one desk. The computer on the desk was purchased by my company but other stuff wasn’t like my mouse or my iPad. I have work Slack on my phone, which is my personal phone. I know I should be, but I’m just not that careful anymore about what I do where.

Granted, I work for a startup. It’s a MBP they had shipped directly from Apple to me. I set it up and configured it myself.

The GitHub Balanced Employee IP Agreement acknowledges that this distinction is arbitrary and unhelpful:

> In California the main difference made by BEIPA is that IP developed with company equipment or relating to the company's business, but in an employee's free time and which the employee is not involved in as an employee, is not owned by the company (but the company does get a non-exclusive and unlimited license if the IP relates to the company's business). This recognizes that from the employee perspective, segregating one's life activities based on ownership of devices at hand or relatedness to an employer's potentially vast range of business that an individual employee is not involved with as an employee imposes significant cognitive overhead and often doesn't happen in practice, whatever agreements state.

- https://github.com/github/balanced-employee-ip-agreement

I hope that more employee agreements move this direction so we can stop trying to enforce this distinction.

Hey this is cool! Looks like a great idea for an underserved market, well executed.

Look up SIBO. Talk to your doctor. Get the breath test.

Interesting. But why can't I filter by city? I'd like to see the percentile plot but restricted to New York. I imagine that the shape and peak of that plot would be different for each city.

The dash is the NOT operator. This is the error that occurs when the query contains only NOT terms. It’s common to check for this special case in query processing and throw an error. Otherwise it’s easy to write a query that returns all or nearly all of the documents in the index.

Try `-1 2`, `-1 -2`, and `1 2`

Linked from the article is the final A/B test they ran: https://www.mediawiki.org/wiki/Page_Previews/2017-18_A/B_Tes...

They had better metrics than just number of API requests

Octopart | New York City | Multiple Positions | Full Time | Onsite Octopart (YC W07) is a search engine for electronic parts. Think Kayak, but facilitating the growth of the Internet of Things (IoT). Every month, 650,000+ electrical engineers and part buyers use Octopart to find parts, research pricing and availability, find datasheets, and select components for new designs.

You'll be part of an entrepreneurial and supportive company whose employees genuinely enjoy working together to overcome interesting challenges.

We use: Linux, Python, Go, MySQL, Elasticsearch, Looker, Redshift, AWS.

Open Positions: Site Reliability Engineer (SRE) * Data Analyst * Data Scientist

Contact: jobs @ octopart.com

Full position details and info to apply: https://octopart.com/jobs

Octopart | New York City | Multiple Positions | Full Time | Onsite

Octopart (YC W07) is a search engine for electronic parts. Think Kayak, but facilitating the growth of the Internet of Things (IoT). Every month, 650,000+ electrical engineers and part buyers use Octopart to find parts, research pricing and availability, find datasheets, and select components for new designs.

You'll be part of an entrepreneurial and supportive company whose employees genuinely enjoy working together to overcome interesting challenges.

We use: Linux, Python, Go, MySQL, Elasticsearch, Looker, Redshift, AWS.

Open Positions: Site Reliability Engineer (SRE) * Data Analyst * Data Scientist

Contact: jobs @ octopart.com

Full position details and info to apply: https://octopart.com/jobs

Octopart | Backend Web Engineer | New York City | Full Time | Onsite

Octopart (YC W07) is a search engine for electronic components. Every month, 500,000+ engineers and part buyers use Octopart to find parts, research pricing and availability, find datasheets, and select components for new designs. You'll be part of a small, supportive, friendly team who genuinely enjoy working together.

We use: Linux, Python, Go, MySQL, Elasticsearch, AWS.

Contact: jobs @ octopart.com

Full position details and info to apply: https://octopart.com/jobs

Octopart | New York City | Full Time | Onsite

Octopart (YC W07) is a search engine for electronic components. Every month, 500,000+ engineers and part buyers use Octopart to find parts, research pricing and availability, find datasheets, and select components for new designs.

You'll be part of a small, supportive, friendly team who genuinely enjoy working together.

We use: Linux, Python, Go, MySQL, Elasticsearch, AWS.

Positions:

* Backend Web Engineer * UX Designer

Contact: jobs @ octopart.com

Full position details and info to apply: https://octopart.com/jobs