aaronds

Neuralk AI | Field Engineer | Full-time | Hybrid (Paris/London)

We are a seed stage startup building Tabular Foundation Models (TFM). This is an emerging space with huge potential.

LLM's like Claude take text as input and produce text. Model's like nano banana take text as input and produce images. Tabular Foundation Models take rows/columns as input, and produce predictions.

The world is sitting on huge amounts of structured data that isn't being used to its potential. Making predictions is hard. Today, most folks are building, training, maintaining their own models. A bit like you'd need to have done back in 2019 with LLM's.

We want to change that. We're looking for an early career Field Engineer to help build out demos, docs, resources, integrations as well as driving product adoption directly with customers. Someone who is passionate about building, has shipped real software (or predictions), has experience with Python and data science fundamentals.

If you're looking for something a bit different than just "hands on keyboard" engineering, and enjoy business as much as tech, this could be for you.

Feel free to reach out at aaron.stillwell@neuralk-ai.com

A bit lower level than most things discussed here but on the topic of overlay networks, I’ve used nebula for years and can recommend it

https://github.com/slackhq/nebula

I run AssetRoom, which sends AI summaries of SEC filings. Found myself wanting a quicker way to discover interesting companies and hear what others think about them.

So I added a weekly poll. I pick a stock, show some key facts, and let people vote and discuss whether it’s overvalued or a hidden gem. Takes a few minutes, usually learn something.

https://www.assetroom.net/diamond-or-dud

By default free plans can run 5x concurrently on self-hosted, 20x minimum for all paying customers, and yes there's a "talk to sales" for >20x on the pricing page

Good to know - and I can see the confusion on that page - I'll pass on the feedback, thanks!

Can you send a link to the page you’re looking at? Thanks!

Pricing should mostly just be users + build minutes (for cloud runners) + storage. There is a few other optional, feature specific costs. Self hosted runners are free, but you need to self host caches/workspaces - our native ones have an egress bill to self hosted runners.

Is that because you have loads of users? (curious CircleCI employee here)

Created a community based poll where we vote and comment on a different stock every week.

https://www.assetroom.net/diamond-or-dud

Would love feedback.

Thanks! Will make a note of this, for sure scope to make them easier to digest.

just seen it- i think you made a typo in your email (.clm instead of .com) - I'll correct and re-send!

I'm working on AssetRoom, a free service to email you noise-free, easy to digest summaries of SEC filings from companies you're interested in.

I often read about interesting public companies (from an investment perspective or otherwise), but fail to then keep up with them over time (sometimes reading many months/years later how successful they were - or not!). I built this to make an easy way for me to follow updates from said companies.

https://www.assetroom.net/

I'm working on AssetRoom, a free service to email you noise-free, easy to digest summaries of SEC filings from companies you're interested in.

https://www.assetroom.net/

Totally. But regardless lots of ideas get funded that turn out not to be "venture scale" and many at an early stage that do not even get to $10m ARR.

That is absolutely absurd that you have built something many have tried and failed to do with millions of dollars of venture capital behind them, all on your own.

Genuine kudos to you, you should be an inspiration to any indie hacker.

Like this? https://circleci.com/docs/2.0/parallelism-faster-jobs/

+1 for iwantmyname - such a simple service - does exactly what I want without any fuss

The S3 SDK allows you to generate a URL that can only be used to upload to a specific path (+ there are other constraints you can enforce, file size / mime type etc)

Correct, I have no need for the original filename in most cases. If I did want this info, for example if I was building a file browser type thing (ala dropbox), then sure I'd keep that in the db.

Personally I'm uploading directly from the browser to S3 using presigned URL's. All files get uploaded to a /tmp directory in my bucket. This bucket is configured so that all files in /tmp are deleted after 1 day (to remove any unsaved uploads). When a form is submitted, I pass the key to the temporary file in the form (via e.g <input type="hidden" name="s3_key">) and create the associated database record. I then move the file from its temporary location to its permanent one upon saving said record.

Feel free to email me to continue this discussion - email address is on my profile.

Something to the effect of my_app/profile_images/org_id/user_id/uuid_thumb.png uuid_large.png uuid_original.png can work.

Then in your app, you can have a way of specifying which version of the file you'd like to reference, for example:

`user.avatar.large -> '<path>/uuid_large.png'`

Not sure if that helps?

Perhaps I'm missing something about your use case, but I only create buckets per application, or sometimes file category (videos, profile images, whatever).

Trivial example following your org/user pattern:

my_app/profile_images/org_id/user_id/aabbccccccc.jpeg

I then obviously have a reference to that file in my db.

Someone please correct me if I'm wrong, but that photo appears to be from some sort of staged shoot, which probably explains why it's not being used as you'd expect.

Edit: Here's another photo that appears to be from the same event http://matrixdaily.com/wp-content/uploads/2016/07/MTX_MAGIC_...

And a photo of the roundabout in a more typical setting: http://www.boostinspiration.com/wp-content/uploads/2011/06/S...

Agreed, and to my knowledge this is a known issue (I'm sure I've seen chatter regarding the problem on the docker forums/github). Nonetheless, with docker-sync I've been able to leverage all the benefits of docker-compose whilest mitigating those slow mounted fs issues, so for now I'm happy with it. At least you can configure it directly into your compose set up, so you don't need to spend time explaining it immediately when onboarding those new developers.

Have you tried https://github.com/EugenMayer/docker-sync ? I've had slow FS issues with docker on mac. Integrating docker-sync doesn't take long and should help you out.

Yes, one option is to do as you've described, for example, use JS to upload the file, then submit a form as normal with a file url / id in the form.

However, you can also create a Lambda that is triggered whenever a file is uploaded to a given bucket and directory. With this, you can react to the file being received and update your database without the need for a second request.

I know this isn't quite the kind of answer you're looking for, but for those who are willing to put even more work in, $150 should be able to get you reasonably far with hobbiest electronics and a 3D printer (that price is obviously not including the cost of purchasing a printer).

Take for example, this open source chassis [1] that is based on the DJI Mavic (retail $1,000). Plenty of fun for those with the time and desire to take this even further.

[1] http://www.thingiverse.com/thing:2004357

What about Cryogenic storage? http://www.bbc.co.uk/news/science-environment-37902773

I haven't looked into it much, but from the little I've seen it looks promising.

Isn't one appeal of the city being able to attract talent from ETH Zurich?