Ask HN: What is your job role and what are the side projects you are working on?

45 points by aryamaan ↗ HN
The question in itself is self-explanatory. Please mention your normal day job role (backend developer, full stack engineer etc) and what are the side projects you are doing currently.

This question is about having insight about what people are doing.

82 comments

[ 3.5 ms ] story [ 159 ms ] thread
I work for a small map tech company and my job role is basically a graduate C# developer, with some configuration duties also to cater our product to different customers. I didn't know C#, only Java, before starting the job so it's been nice to learn about concepts like WPF and Linq this past year.

I do some freelance work for a small business owner in which I code small Java apps that hook to the eBay and Amazon APIs, for the benefit of his business. It makes me a little money but I've recently asked for a pay rise.

Other than that, I want to make my own mobile apps using tools like Xamarin but I haven't really built up the confidence in myself to create any of my ideas. They mostly seem like things that have been done before or where I don't see that they might get as much use as the idea does in my head.

On top of that, I enjoy my free time. I find it very hard to come home and have the energy to code after my 9 to 5 where I've been doing the same thing, staring at a screen, 7 and a half hours. Instead, I try to exercise a fair bit and build hobbies outside of programming. I'm a bit of an everythingist, admittedly, though it's not a quality I think leads to becoming a master in programming.

Has anyone else faced similar problems to me? What is your answer to your own question, OP?

I have the same thing with coming home from work and just being done with screens. What's worked for me is coding for 1-3 hours before work, and treating one weekend day as a normal workday but for myself.
I've sort of tried the second one and usually do get a bit done at weekends. I like the first one though, I'll try that too. It's all a question of whether I can get up that early!
My job title is just Developer, but I work primarily in a front end capacity and currently I'm working on a passion project, a cross platform mobile app that I'm transitioning from Ionic to native, a Swift project that's a pretty simple voice note recorder that starts recording when you open the app, and a teaching/building project with a friends art collective, slowly building it while teaching them how to code each part.
I'm a Mechanical Design Engineer in normal day job and I conceive spectrometers for LULI laboratory. The side projects I'm currently doing is my Startup Air Spectral (http://www.airspectral.com), where I'm developping a miniature hyperspectrale camera. I'm doing currently another project where I'm not the leader. It consist on builting a new motor concept based on the cogeneration principle. In parallele of my Mechanical education, I learnt programming and lead with my friend during a moment the project lylconcepts (http://lylconcepts.com/)
Software developer; Working on a tool to do faceted search across google docs / dropbox / onedrive
Job role: I'm the manager and lead dev of a team working on a number of military software projects for the US Air Force.

Side projects: I've been tinkering with new approaches for doing MVC and templating in JS. See my projects on GitHub: https://github.com/kethinov

Web Developer; building (or trying to) a roguelike in my spare time.
I'm a product manager. Side project is building a genome data ingestion pipeline with my dad. We're both really passionate about giving people a faster and cheaper solution to processing genome data.

Current project is to allow researchers to send fully sequenced genomes to our server, align the sequenced genome to a referenced genome, then store the data in a performant way that can be queried against. Right now, its either too technical for the researchers or too expensive. I have dreams of making it something big but time is a limited resource as well as finding people who A) know how to build massive data platforms at scale and B) know enough about genomics to build the platform.

We're coming along well so far. The alignment of sequenced genomes to referenced genomes if done (although not scale ready yet). Currently working on/learning the data storage side and what to do there.

Shameless plug: if anyones interested in our project, we could use some help.

Sounds interesting.
Thanks. We're pretty excited about it. There's lots of big challenges to solve, especially on the data from. Full 30x coverage of a human genome is (uncompressed) around 180gb. So the core problem we're trying to solve right now is: 1) Compressing the files locally and sending via FTP to server 2) Aligning 180gb files against the reference genome (180gb also) in less than 5 minutes 3) Storing that data in a performant way so that you can do comparative genomics (compare multiple genomes against each other) 4) Do the above with 100's-1,000's of genomes a day. It's pretty rewarding. Apart from the cool data side, it's feels good to know that building something like this could really help accelerate critical research in finding diseases.
Is anyone using your pipeline? There are lots of fully developed commercial pipelines for human WGS (Illumina BaseSpace, DNAnexus, Seven Bridges, etc.) Depending on the hardware you have available bowtie2 might not be the best aligner. What are you using for variant calling? If you're using the GATK Haplotype caller you're going to have issues with commercial licensing.
No one is using it yet. Still just a side project :)

Seven Bridge & DNAnexus probably have the most complete/thorough system (imo). The challenge though is doing it fast, at scale, and user friendly enough. Seven Bridges is cost effective and can definitely scale but it's not necessarily the fastest solution on the market. While these guys offer the full end-to-end solution, we're just trying to focus specifically on accelerating the pipeline. Maybe in the future we'll get to the full circle, but that's not our focus.

We use a variation of bowtie2 that allows us to scale well. We're able to align a 30x genome right now in about 8 minutes and are doing a couple more tests that might get us down into the <5 min range. The goal is to be able to do multiple at once in <5min with our portion of the pipeline taking less than 10 min total.

There are universities, such as Harvard, that are trying to align 1,000s of these a month but the current providers can't keep up. So, we're seeing if we can provide something that can help them out.

Don't worry, I work for a company in the space and fully understand the challenges :) I have a really hard time believing that you can align a 30x genome in 8 minutes because the I/O time is longer than that, and bowtie doesn't use one of the faster algorithms for alignment (usually the kmer hashing methods are faster when you have machines with enough memory. Microsoft's snap and Illumina's isaac are two examples). What kind of hardware are you benchmarking on?
Check out NVBowtie. We don't use standard CPUs to do the alignment. The benchmarks were ran on 2 Nvidia K80s which were in the same blade. To scale it, you need Infiniband between the blades (found this out the hard way). The genomes are loaded into memory to reduce read times.

Just to be clear, we're not reading this directly off of the sequencers. Our assumption is the sequenced data is already stored in which we load the data onto the cluster.

I'm not necessarily the technical one of our group unfortunately but, if you're interested, I'd be interested in picking your brain.

Ah, gpus are cheating... :) What I meant about i/o is that if you're reading the data off of disc your read/write time is probably longer than 8 minutes. With SSDs you can obviously go faster. One problem with scaling is that you need a bunch of machines that have the gpus available (although if you can get the whole pipeline under 30min you wouldn't need many machines). The cloud genomics companies mostly use AWS and Google cloud for their compute and I don't know what sorts of non CPU compute resources are available. You would probably be interested in looking into Edico Genomics' Dragen FPGA. Is there a way to PM on HN?
Out of curiosity, what genome assembler are you using for your project? (I'm having to build a genome assembly pipeline with a reference assembly as well for a bioinformatics class I'm auditing at a local college, not a professional in this field either). The papers I am reading and trying to replicate uses SMALT.
Assuming you're referring to the alignment side of things. We use a variation the Bowtie2 algorithm that allows us to align multiple genomes at once to the same reference genome.
Thanks for your reply, rfc. In my Bioinformatics class, we went through Bowtie algorithm funny enough last week (the vague details I still remember are the funny way it compresses fragments as rotations and then goes onto transform the rotations). Gl on your project.
Software developer, focusing on web service and application development.

Side projects, a few gigs on Upwork.

I don't have a title, we are a pretty small company. I manage the systems, networks and our virtual environment. On the side I enjoy projects that don't include a keyboard or mouse like fixing old mopeds and dirtbikes. I'm currently restoring a 1996 Tomos Sprint moped I had when I was a kid.
In my day job, I'm a full-stack developer at an ad agency.

Side Projects:

Early next week, I'm going to be releasing an Android word game.

Revenue-producing side-projects: https://SendToMyCloud.com & https://PrivateForms.com

Other side projects: http://artpacks.org, http://aboutbrowser.com & http://norefer.link

Your revenue-producing side-projects are strangely inspiring. If you don't mind me asking, how do you market them?
Mostly, I don't.

In the beginning (for each) I found a few people asking on forums for software that solves the exact problem that these products solved, and then posted the software as a solution.

Since the first few weeks, I haven't done any marketing at all—everything comes in via search, or referral traffic.

That's cool! Thank you for sharing.
What stack do you use for your projects?
It varies a bit from project to project, but I usually start with a custom PHP MVC framework: GluePHP for routing, Savant3 for templates, and PHP-ActiveRecord as an ORM. I'm definitely someone who prefers to work with small, understandable modules with very limited scopes/responsibilities.

I've recently considered polishing up the most-used bits and packaging it as a proper framework, but I probably need a kick in the ass (or even just a hint of interest) to do so. :)

Artpacks.org strays from the framework a bit by using Backbone.js for the front-end. So does the game that I'm working on.

Full stack Software-Engineer. Side project is a Toolkit for high productivity web-development in Go language built around the concept of code generation [1]. I'm using it for another side project (online start-up incubator platform).

[1]: https://github.com/colegion/goal

Nominally software engineer, doing Devops but moving to a backend position. My main side project right now is building a software synthesiser in Rust
Javascript developer, mostly doing front end work and some node server side and for build scripts. Side projects (I've been trying to cut these down) are: 1) helping my brother with marketing a food catering business, 2) developing my product management skills 3) building a local food e-commerce site.

Before you say it, yes, I have a problem focusing.

day job - full stack developer for a local company that develops worksheets for k-8 teachers

side job - full stack developer for a website that provides continued education for dental students and dentists

I'll generally take the random freelance job that comes around, both remote or local - building a Wordpress website for a local plumbing company this week.

If those 3 aren't consuming all of my time, I've been planning an idea for a programmers network, but i haven't had much free time lately :(

My day job: front-end developer, which means different things to different people. Mainly do JS work right now with Angular JS on a large enterprise transactional application.

My side project: Working to create an agency that helps a specific niche business. We already have close to a dozen clients on a monthly subscription that are generating a good chunk of revenue. A few more clients and I can forgo my current contracting gigs and do it full time.

I teach high school math and science.

I've been writing in the mornings and late at night, and I just finished an intro Python book: http://nostarchpress.com/pythoncrashcourse It will be out next month, and I'm very happy to focus on supporting the book now that the writing is finished. I love writing at the introductory level, where the focus is half on technical issues and half on helping people learn to think as a developer.

Writing a book has significantly increased the degree of professionalism I bring to my side projects. My school district just adopted a new learning management system, but there's no functionality included for generating report cards, so I'm working on a project that automates the process of generating report cards. The software is relatively easy and enjoyable to write, but the politics involved in making sure this work is implemented correctly into the overall district processes is challenging. Having worked through all the details involved in publishing a book helps me identify and work through the many non-technical issues involved in fully implementing a side project.

Day job: Build Engineer in games Side project: browser based snail simulation.
I'm a Software Engineer at my company. I used to work on a large Rails project, but now I am on a team where we are building out a smaller product with a React/Flux frontend. I work mainly on frontend stuff right now.

I have a small side project for a family member. He's a commercial beekeeper and needed a way to track information about his hives, so I am making him a mobile app (built with Ionic) that lets him put QR codes on the hives and track everything he needs about them.

My wife and I just had our first baby, and I'm starting Georgia Tech's OMSCS program in the Spring, so I don't have much spare time right now. But off and on I have been learning Clojure/ClojureScript, Haskell, and more about React and its ecosystem. I also am teaching an introductory JavaScript class for a local tech "bootcamp" company, and enjoying it a lot.

Right now, I am primarily a front end web developer.

I always have a lot of different projects running around in my head. I have two big ones right now. The first is a React Native app integrating with a Buddypress site. The second that I am working on is a plugin that integrates a popular API with Buddypress that I am going to post on CodeCanyon.