Globality (https://www.globality.com): Menlo Park | Palo Alto | London | Tel Aviv | Full-Time
Globality builds an online marketplace for companies to buy and sell professional services (think marketing, consulting, legal, HR, and IT).
We enable talented small businesses to bid on the same projects as multinational corporations, allowing for richer competition and an even playing field in world commerce.
At the heart of the network is our ML recommendation engine that surfaces the best-fit providers for each given project.
There are open positions for senior engineers across our FE, BE, SRE, and AI organizations.
We work primarily with python and JS (Flask, GraphQL, express), we have everything as infrastructure-as-code and we have ~200 micro-services running in production.
On the AI side - we focus on core research for NLP processing and developing recommender systems. If DL transformer networks, reinforcement learning, and large scale web crawling interests you then get in touch!
I've never done it for myself. I did, however, do it for screencasting (Twitch and YouTube).
There are a couple of things here:
1. Talking to yourself is extremely useful. You'll be surprised by how effective it is.
2. A recording is not very useful because discovery is awful.
Even though all my content is on YouTube, There's no way I can find that one time I fixed an interesting bug or implemented something complicated.
Without proper logging of what you did and indexing in the video, it will just be like any other video you take with your GoPro, hours of nothing.
What I found to be really effective with my teams is this:
1. Draw your thoughts.
2. Write up a mini-plan of how you want to "attack" this
3. Talk through it, even if it's only to yourself.
4. Pseudo Code and Pseudo Flows of data
Interesting... Voice transcription software can close caption and you can search by that. You can have keywords you can use for searching. TODO, BUG FOUND, BUG FIXED, FEATURE X. Perhaps the tech is not yet solid for searching, but I feel that the basic elements is there. I'm more interested in doing this for solo project, but this can be interesting for teams too, what if you can look through a git history and match it up with the video of when it was developed and hear the author's line of thought? The tough part of it of course will not be the tech but the behavior of getting folks to speak up and dump their thinking verbally.
Globality builds an online marketplace for companies to buy and sell professional services (think marketing, consulting, legal, HR, and IT).
We enable talented small businesses bid on the same projects as multinational corporations, allowing for richer competition and an even playing field in world commerce.
At the heart of the network is our ML recommendation engine that surfaces the best-fit providers for each given project.
There are open positions for senior engineers across our FE, BE, SRE, and AI organizations.
We work primarily with python and JS (Flask, GraphQL, express), we have everything as infrastructure-as-code and we have ~200 micro-services running in production.
On the AI side - we focus on core research for NLP processing and developing recommender systems. If DL transformer networks, reinforcement learning, and large scale web crawling interests you then get in touch!
I realize my comment hit a sensitive spot but it's really not meant to be negative.
It is 100% legit to give that up, it is 100% legit to go with the clear winner in your mind (it is the same in my mind too), but the technical reasoning is weak IMHO.
if you wanted to support both, you can, but you clearly don't (and that's ok too).
It's great that the simplicity argument doesn't really apply. It's a testament to developers getting better at programming in the large over the years. The database layer can be neatly abstracted away from the business rules.
We are very heavy users of terraform. We have thousands of lines of HCL and our own providers etc...
Terraform 0.x before 0.12 was heavily limited due to the syntax, you just worked around so many limitations. We are actively ignoring those limitations and we started working without our own templates using jinja that just generate vanilla terraform. No counts and no if/else needed.
Maybe with .12 we can move some of these back to plain terraform.
Upgrade so far is not smooth though, there are a lot of pains with the type system vs the plain old “I’ll figure it out for you”.
ALL of this being said, I really appreciate Hashicorp’s work on this, we could not imagine our life without terraform.
