So, it was hard to us to find some job for the guy but we were open sourcing our job. I told the guy to work on a next version of our open-sourced Python package.
After we released it, it went #1 on Hacker News. I put him everywhere in the credits. You can still see him in our “Team” section.
The next day he got 2 offers.
It got me to the interview stage often and then it was up to whether I knew the nonsense whiteboard leetcode situation, which I've never been particularly good at.
If I got the job it was because they were pre-sold on step one's effectiveness I've always felt like.
I didn't study machine learning per say, I followed dozens of online courses but no diploma in this exact field, and my last job was in traditional software engineering.
Despite this, I managed to do two small projects (one of music genre recognition  on my own iTunes library, and one of eye motion recognition ). For both I had published the code on my GitHub, and written a small Medium article to expose the problem I was trying to solve, the data available, the issues I faced and how I overcame them.
I personally believe that this is all the recruiter wanted to know, and why I encourage interns and candidates to work on their side projects.
References of the projects:
> If you are going to find a project, I encourage you to find something particularly small and particularly durable (which is to say, does not require constant or near-constant attention), so that you give yourself the ability to take a break from the project for weeks or months while it accumulates interest.
*edited for formatting
I wanted to make the jump from system administration to devops engineering. As i started started learning ansible i noticed that something was a little off and submitted a small pull request about one of the core ansible modules.
That didn't get very far because the maintainers didn't like the way it was and also i hadn't time to fix it the way the maintainers wanted it to be (not having time was also one of the reasons for me wanting to make the jump).
Despite this, I included a link to that pull request in my cv anyway (along with other OSS contributions) and that definitely helped, in two ways:
1. during the technical interview i was asked about that pull request and that meant i was able to direct the discourse where i wanted and where it was most favorable to me and basically talk about stuff i knew about (i mean, i did the stuff).
2. as i have been told much later, their gut feeling when taking the decision about my hiring was that i wouldn't have had much problem in figuring how stuff work even though i wasn't technically qualified for that position (in my opinion I was like ~30% qualified). Worst case scenario, i could have been let go during the probation period (I was not).
That was my experience.
Needless to say, that helped me getting hired. I then had to work my a.. off to fill the gaps in my competencies and learn a lot of the new tools. It's been almost a year now and I'm fairly confident with most of the stack, even though there are a few components that I still have to properly study and learn.
So far, it's been a pleasure.
A few of the posts were research based and therefore they wanted to know more about my insight. A few of them were tutorials and they wanted me to build something for them. Hope this helps. If you have any other question, send me an email. (in profile)
My mantra to everyone is to just build things. Don't wait to be asked. Build things. Makers make things.
I built an ecommerce site out of curiosity using some of the lesser documented PayPal APIs and services for micro-transactions (even talking to their developers to get concrete answers.) While relaying this to someone at a tech meetup I was offered a job on the spot from a fintech company CTO because he'd been trying to solve the same problem and was impressed with how far I went to solve it out of curiosity.
Another time I got invited to beta a product from a company and was able to show them a variety of uses cases by building small projects around it, which lead to a job offer.
I wasn't doing particularly sophisticated coding, but I knew enough to build everything from mobile apps to server-sided apis.
I never would have made it past the recruiter for either company had I taken the traditional route because I've never formally worked for anyone in tech. However my habit of learning all kinds of frameworks and outside thinking was something they realized they wanted.
In the latter instance I also blogged about their technology which helped showcase its potential.
On the 3rd interview I asked to look at the CV that they'd been given by the recruiter. It turns out the recruiter had deleted both my blog and my GitHub link.
Presumably the recruiter feared the company might be able to contact me directly via those links. Or maybe just didn't understand what they were.
A classic case of the recruiter looking after their own interests instead of those of their customers.
So - in conclusion - the last time I searched for a job - my blog with > 250 posts  had zero influence on the outcome of my job search :-)
It's a good thing I wrote the posts for the fun of it, and not with the intention of improving my job chances.
Almost every professional job I've had since the 1990s was because I liked Slashdot in 1999 and wanted a site just like that. So, thanks CmndrTaco!
It was based off of class times, so when more classes were in session near the buildings of those parking lots, the color would change.
I put it on my resume, and had it in the Google Play Store during my time at university. Got asked about it during interviews and every interviewer really liked the fact that I saw a problem, proposed a solution/assistance, and published it to the world.
* blogging (http://semanticbible.com/blogos/, now rather moribund)
* creating hobby projects (like The Composite Gospel Index: http://semanticbible.com/cgi/cgi-overview.html and New Testament Names: http://semanticbible.com/ntn/ntn-overview.html)
* submitting talk proposals to biblical studies conferences, though I had virtually no formal education in that field.
It was a classic "FieldA + FieldB = Innovation" strategy: but at the time it wasn't a deliberate strategy, just what interested me. I met some folks at the conferences who made a software package I'd never heard of (Logos Bible Software), which led to discussions and a job offer. 13 years later, i'm the Director of Content Innovation there. My side projects were responsible for the original connection, as well as demonstrating my interests and competence, and the novel contributions I could make to the company.
Another one, I remember was a blog I started to experiment with Google Adsense. It didn't make much money, but once or twice a day made a few cents. There were some book reviews, and links to "read more" on a few article (for readers who wanted to research more on the subject). I added Amazon referral links to it. Was shocked a week later to see that I had made around $50 or so with one click because somebody who had clicked the link bought some expensive item on Amazon! While the blog was alive, Amazon ads / links made me more money than Google Adsense. That was some fun experiment, while it lasted.
It helped to show case that I understood several parts of the software development cycle, my communication style and documentation. This, I believe, would be hard to describe on a CV. 
It also helped me a lot to practice and learn things otherwise I wouldn't have, and that I put on practice for the interview task. 
Here are my examples:
Since I'm not working in that exact field anymore this blog has lost its rankings. I do have another blog that I'm building (https://integrated.ee) which I kind of hope achieving something similar in the future. So far from the new blog there has been one contact which didn't lead to any significant co-operation.
Because I had something (even though it was crap), I was able to talk my way into being the cofounder at a startup, then a job at a FAANG at a crazy salary. The FAANG recruiter mentioned specifically that they saw I was running my own shop (even though I never landed any accounts > $10k in the 6 months).
Didn't do any leetcoding, just keep on writing helpful tutorial