I've recently become curious about the scene, and discovered quite a bit going on. It seems most tech coops are operating more like agencies, taking on client work. Haven't seen many developing products.
Here is a new but somewhat active matrix channel where some tech coops folks hangout: https://matrix.to/#/#techcoop:autonomic.zone We plan to have an informal call to chat there on 22nd of February at 17:00-18:30 UTC.
Note that a big problem with building a product tech coop is that you exactly can't as easily raise millions: investors by definition cannot take stake in a coop. Instead, you need a business loan, and its less common to find financiers willing to loan large rounds for high risk investment. That may be a cultural thing though, because in theory, sufficient interest could offset the risk in aggregate I guess.
It might interesting to set up an open source group that provides an ecosystem of re-usable tools and libraries that interoperate well, and solve problems that are otherwise solved with with repetitive flakey internal implementations by several players in an industry.
Sort of like Apache Software Foundation, Spring Framework (for java), Boost (for c++), but higher level and industry-specific, e.g. for finance, some libraries and tools for handing FIX protocol, parsing file formats from market data vendors, also tooling for change management / releases etc. I am sure each industry has its collection of common use cases, like geology tools/libraries for mining, or handling genetic data for bioinformatics/pharma, etc.
The group, though, should have a corporate-friendly outward branding, be composed of people actually working in the industry, and somewhat selective with membership. A goal, perhaps not mentioned on the website but known to members, would be that the organization exists for the benefit of its members getting hired in the industry, consulting work, hiring each other when practical, books, training, etc. on using the libraries and tools.
In a way, it's really not anything too original, and I am sure close to how many open source groups have always operated in practice, crossed with a something like a guild/union/professional network. The main thought would be to do this for what seems like a "missing middle" of industry-specific stuff where there are maybe tens or hundreds of potential client companies, so not totally generic like kernels/compilers/spring/boost, but not a big enough market to attract a VC backed company, but big enough to offer cushy jobs for a few dozen people who would never have to do leetcode interviews ever again.
You mean your standard consulting company that works on a project basis? This isn’t anything new.
I worked for a startup founded by two non technical brothers who brought in an outside consulting company to do all of the work and then they brought in a CTO to bring it in house and I was one of his first technical hires.
You’ll be competing with cheap outsourced developers where only the tech leads and project managers are paid American wages.
You are basically doing staff augmentation which is the fifth level of hell
No matter how you frame it, you are still competing with the “Americans who are the leads and people from India doing the work for less than Americans” companies.
Google, Facebook, Instagram, Stripe, Whatsapp, Coinbase, PIntrest, Amazon, Dropbox did well did they have a couple of team leads in US and rest in India at the start?
(To be clear: nothing against Indian talent. My point is you don't need to compete with global outsourcing unless the value proposition isn't high so you need cheaper devs)
They aren’t hiring the cheapest developers who can barely get into one of the WITCH companies where they are competing for contracts by cheap American companies.
You are always competing with global talent - especially if they will work for 25% of your cost and these days can just be LLM supervisors.
My biggest issue with Indian developers in India that were contractors (not people in the US) is they never take the initiative and “work to rule” and will not speak up if they think there is a better way.
I don’t see that in my current LatAm coworkers. They have no issues speaking up, taking initiative, dealing with ambiguous requirements, etc.
There are many private companies that operate under this model, and they’re usually small to medium sized (<250 people). The success of these companies often depends on their area of expertise and leadership.
Unfortunately what is a “specialization” these days that broadly applicable enough that it isn’t a commodity? Front end? back end? Mobile? “DevOps” [sic]? AWS?
I've recently become curious about the scene, and discovered quite a bit going on. It seems most tech coops are operating more like agencies, taking on client work. Haven't seen many developing products.
Here is a great list of tech coops: https://tech-coops.xyz/
Here is a new but somewhat active matrix channel where some tech coops folks hangout: https://matrix.to/#/#techcoop:autonomic.zone We plan to have an informal call to chat there on 22nd of February at 17:00-18:30 UTC.
Here is a Discourse forum for tech coops: https://community.coops.tech/
Note that a big problem with building a product tech coop is that you exactly can't as easily raise millions: investors by definition cannot take stake in a coop. Instead, you need a business loan, and its less common to find financiers willing to loan large rounds for high risk investment. That may be a cultural thing though, because in theory, sufficient interest could offset the risk in aggregate I guess.
A recent thought:
It might interesting to set up an open source group that provides an ecosystem of re-usable tools and libraries that interoperate well, and solve problems that are otherwise solved with with repetitive flakey internal implementations by several players in an industry.
Sort of like Apache Software Foundation, Spring Framework (for java), Boost (for c++), but higher level and industry-specific, e.g. for finance, some libraries and tools for handing FIX protocol, parsing file formats from market data vendors, also tooling for change management / releases etc. I am sure each industry has its collection of common use cases, like geology tools/libraries for mining, or handling genetic data for bioinformatics/pharma, etc.
The group, though, should have a corporate-friendly outward branding, be composed of people actually working in the industry, and somewhat selective with membership. A goal, perhaps not mentioned on the website but known to members, would be that the organization exists for the benefit of its members getting hired in the industry, consulting work, hiring each other when practical, books, training, etc. on using the libraries and tools.
In a way, it's really not anything too original, and I am sure close to how many open source groups have always operated in practice, crossed with a something like a guild/union/professional network. The main thought would be to do this for what seems like a "missing middle" of industry-specific stuff where there are maybe tens or hundreds of potential client companies, so not totally generic like kernels/compilers/spring/boost, but not a big enough market to attract a VC backed company, but big enough to offer cushy jobs for a few dozen people who would never have to do leetcode interviews ever again.
You mean your standard consulting company that works on a project basis? This isn’t anything new. I worked for a startup founded by two non technical brothers who brought in an outside consulting company to do all of the work and then they brought in a CTO to bring it in house and I was one of his first technical hires.
You’ll be competing with cheap outsourced developers where only the tech leads and project managers are paid American wages.
You are basically doing staff augmentation which is the fifth level of hell
Sounds like he is talking about a startup per-se where say 10 devs get 10% ea?
No matter how you frame it, you are still competing with the “Americans who are the leads and people from India doing the work for less than Americans” companies.
Google, Facebook, Instagram, Stripe, Whatsapp, Coinbase, PIntrest, Amazon, Dropbox did well did they have a couple of team leads in US and rest in India at the start?
(To be clear: nothing against Indian talent. My point is you don't need to compete with global outsourcing unless the value proposition isn't high so you need cheaper devs)
They aren’t hiring the cheapest developers who can barely get into one of the WITCH companies where they are competing for contracts by cheap American companies.
You are always competing with global talent - especially if they will work for 25% of your cost and these days can just be LLM supervisors.
My biggest issue with Indian developers in India that were contractors (not people in the US) is they never take the initiative and “work to rule” and will not speak up if they think there is a better way.
I don’t see that in my current LatAm coworkers. They have no issues speaking up, taking initiative, dealing with ambiguous requirements, etc.
The one I’ve heard of is Igalia
https://www.igalia.com/
There are many private companies that operate under this model, and they’re usually small to medium sized (<250 people). The success of these companies often depends on their area of expertise and leadership.
And how cheaply they can compete…
Or what specialized niche they address
Unfortunately what is a “specialization” these days that broadly applicable enough that it isn’t a commodity? Front end? back end? Mobile? “DevOps” [sic]? AWS?
None of those are.
"Broadly applicable" is always in tension with "niche".