xstartup

No frameworks. We use http://www.gorillatoolkit.org/pkg/mux and sometimes we need a few more libraries for GWT/ElasticSearch/PostGres. I know there more performant options but we've found no reason to change so far.

So you said you're deploying 5000+ web apps a quarter. Are those new web apps?

Yes, 70-80% of them are new apps. Small apps, not multiple month project, mostly CRUD. Much easier/faster and cost-effective for us to develop what we've already done. OId app usually requires adding few routes and shuffling around the frontend elements or adding some small CSS tweaks. We use a simple list to accept/reject projects. We never undertake any complex project.

How many web apps are you deploying per dev head count?

These are small apps. We've 1500 developers in a very low-cost jurisdiction.

Here are the main benefits of patents:

1. Helps in IPR negotiations. When multiple startups are competing in same space and poaching same employees, highly likely your tech will end up being similar to your competitor. Maybe some of the processes are exact copies. Then if you've no patents and you get sued, good luck defending yourself.

We solve it by creating landmines of patents around our actual application. So, that if we end up infringing someone else's patent, we'll be able to find some of their infringement then negotiate the settlement.

If you've nothing, you can't defend and your settlement will not be something in your favor.

Conversation with competitor: Mate! you violated my patent X and I violated your patent Y, we'll get nothing out of the legal battle, the damage is of the same amount. Why not stay quiet and keep the lights on? Who doesn't love peace?

2. Patents in employee's name often boost their confidence (if not their wealth). So, you get some productivity boost from their motivation. Sometimes, it's like keeping a score for them and we payout enough to host a memorable dinner with their family.

Conversation with Self: You know like I've 15 patents under my belt already and 5 more pending. I am an expert already I think. Everyone will be proud of me.

3. Keeping new entrants out.

Conversation with Self: Oh, it will be expensive to navigate this patent landmine, so let's choose some other niche/industry for my new startup idea.

We exclusively use AWS Lambda, DynamoDB or if the app is complex, we use Amazon RDS.

For search - AWS Elasticsearch

Our preferred stack is Go + VueJS.

WHY GO?

It's much easier for our senior developers to ramp up a team of 10 developers on Go than say Python or Ruby. Suprinsgly, we have a few guys who are Go experts, they unblock all other developers when it comes to best practices or some weird bugs.

WHY VUEJS?

It's has been a lot easier for us to introduce this to our frontend developers than any other thing.

WHY NOT RAILS?

We found it difficult to make SPA apps in Rails and ramp period for developers is much higher than Go + Vuejs.

WHY NOT PYTHON?

I loved Flask but we really need Type checking, we are not able to sleep without it. And Go has enough of what we need. Yes, we've briefly experimented with Haskell, OCaml, Scala etc.... but way difficult for new devs.

WHY NOT JAVASCRIPT BACKEND?

We found it much easier to write backend code in Go.

IDE/DEBUGGER

I am a pro-VIM user myself but it's so refreshing and fun to watch new developers speed up on Gogland IDE and the debugger makes developing in Go much easier. This is something only old languages like Java or C# has. It's way more difficult for new developers to pick up Java or C# in our experience. They've way more features which our developers found confusing. So, Go stayed everything else was thrown out including Haskell.

VUE CLI

We use https://github.com/vuejs/vue-cli No it does not require deep knowledge of webpack. There are just a few guys on our team who knows webpack/vuejs/javascript inside out in extensive details, rest of the developers simply use their custom designed/documented template and it satisfies 100% of our frontend needs. They unblock everyone on our team if any developer gets stuck with webpack.

DEVELOPERS ARE OUR BIGGEST COST

Keep in mind, we do not have access to the best developers at the budget price.

TRAINING

We give new developers as much help and training as they need. Our developers can call the pro guys anytime (even on weekends) and our senior developers love the respect/rewards they get in return.

COST

AppEngine costs $30 a pop, for 90% apps which receive less than 100 hits a day, it's expensive. This is where Lambda wins.

CI

Multi-tenant CI is still a pain in 2018. Current workflow with Gitlab sucks. We wish we could have a simple file in our repo and make 1 curl call to run the CI without any initial setup. But it will eventually happen.

PAYMENT PROCESSING

If the SaaS product needs multiple payment methods like Stripe, Braintree, PayPal Bitcoins, Ethereum, Wire Transfer we use http://killbill.io/ which is multi-tenant, opensource and self-hosted. Previously, we were using WHMCS, Amember pro, and Chargify but it ended up costing way more and was a lot more fragile at the end, completely failing monkey wrench test and our new developers found it much harder to wrap their head around 3 different products. A customer has to simply enter their API credentials for the respective gateway (Stripe or PayBear or BitPay) and we integrate it in minutes.

BIGDATA

If we need complex large aggregation, we haul in Athena or BigQuery.

Our developers prefer BigQuery but Athena is sometimes cheaper in the long run and makes it way easier to import/partition your big data.

STREAM PROCESSING

If we need anomaly detection or preprocessing of statistics we use Kinesis streams.

Because of this setup, we primary run on an army of bootcamp graduates and we are able to provide really cheap service to non-profits and other companies who are not able to afford even a dedicated freelancer.

FORGET VALUE, WE WANT TO BE CHEAP

"Cheap" is a wrong word to use as a marketer/sales guy. But yea, since I am now in technical position, so I remove sales guy hate and wear tech hat, making our creation as cheap, robust and secure as possible while still satisfying the requirements.

WHERE IS THE FUN

We drive our major profits from AdTech business (which involves stressful high tech stuff but it's lot more fun) and consultancy is just to validate our solutions across a large number of our portfolio customers. We've to ensure, their setup is as cheap, robust and secure as possible.

IMPROVEMENTS

As I want to improve my strategies and methods as much as possible, feel free to challenge my choice. It should not just work for 1 developer but an army of developers and it should make it trivial to add more developers.

This is something, I never saw in the US.

https://www.youtube.com/channel/UChXICgGipKvJbtzKfM1SNoQ

It supports many payment plugins:

stripe 4.1.1

braintree_blue 0.3.1

paypal 5.0.9

adyen 0.5.10

avatax 0.4.1

email-notifications 0.3.1

accertify 0.3.0

analytics 4.2.5

and many more....

stripe 4.1.1

braintree_blue 0.3.1

paypal 5.0.9

adyen 0.5.10

avatax 0.4.1

email-notifications 0.3.1

accertify 0.3.0

analytics 4.2.5

currency 3.0.0

cybersource 5.2.5

dwolla 0.1.0

firstdata_e4 0.2.0

forte 0.3.0

kpm 1.1.2

litle 4.0.0

logging 4.0.0

orbital 0.1.10

payment-retries 0.2.4

payu_latam 0.3.0

payment-test 4.3.0

securenet 0.2.0

zendesk 3.0.1

payment-test 4.3.0

securenet 0.2.0

zendesk 3.0.1

You can even roll your own payment plugin: http://docs.killbill.io/latest/payment_plugin.html

Here are the screencasts for your business team: https://www.youtube.com/channel/UChXICgGipKvJbtzKfM1SNoQ

Most of it might mean absolutely nothing to a customer. But you think the price you charge is fair.

Don't conflate the efforts you put, with that actual value of your product.

Many software might have the same utility as of a diamond and may have the same price as a diamond.

Why should I care if that makes me look intellectually weak and poor imho? I am from old banker family, I love diamonds and I collect them. I gave you other reasons for which I collect all these diamonds.

If you ask me in an interview, I obviously won't mention these views and give you a politically correct answer. But this is the internet, I am using anonymity, where is the problem with that?

I am just sharing my views, I don't care about your judgment.

> But here's one thing that "fuck you money" won't buy: class.

I am born into it, there is no reason for me to buy.

In a hypothetical world, if a diamond could be bought for $10K and sold for $10K, where is the risk? Then it's no longer for only rich, others can take a loan from a bank and purchase diamonds.

When I pay for a diamond, I don't pay for its intrinsic value. I pay for diamonds because other people around me feel they are expensive and you've to be rich enough to set your money on fire (like you do while purchasing a diamond). It's a signal for others that I am "fuck you money" rich.

The very same reason for which I purchase a sports car, yes a sports car does have some advanced engineering but the resale value is not same.

As long as diamond purchase price stays up I'll keep buying. Only thing I want is that no other people are able to acquire them for cheap when I am willing to pay a premium. (I hate price discrimination, i've also worked hard for my money, why should anyone get them cheap?)

Please stop treating me like a victim of De Beers campaign.

Through which you can sync your user database to WHMCS and account status like (active, suspended, terminated) etc...

But if you've stripe payment plugin (or any other plugin for that matter) enabled in WHMCS, a customer has to login to WHMCS and make payment in WHMCS only. So, it gets in the way.

The approach to integrating SaaS app with WHMCS is quite hacky and involves pain.

And many things are designed with hosting industry in mind.

For accounting, we simply use WHMCS reporting, see: https://docs.whmcs.com/Reports We extract the various reports, send it to our bookkeeper and that's it.

The billing logic is pretty easy:

1. If the subscriber has EU IP or the billing address is of any EU country, add VAT per their country's rate to their invoice if they do not provide a valid VAT ID.

2. We use https://vatlayer.com/ for VAT rates, geolocation using cloudflare ip gelocation https://support.cloudflare.com/hc/en-us/articles/200168236-W...

3. Currently, we use WHMCS for subscription management which costs $39.95 per month. Why? because it supports many payment methods! that's very important to our business. It also takes care of VAT etc....

Or if you are selling bitpay and other plugins for killbill, i'll gladly pay more.

We added web money, cryptocurrencies, wire transfer.

And it has increased our conversion rate by 10%, we are in B2B market internationally.

Look into: http://killbill.io/ for SaaS subscription management. It's free and opensource.

Killbill seems interesting. But I am unable to add bitcoin plugin in killbill. It seems to be 99% of what I am looking for.

But still this place does need user friendly solution.

2. There is license requirement, taxi medallion like setup with lawyers, doctors which keeps overall supply low.

3. It's just not just your salary which company pays. There is a cost to company per employee if the ratio of your salary: cost to company declines, you are a problem for the company and company wants to invest in an employee with a better ratio.

Also, you are making an assumption that we can look at an employee's productivity in isolation.

No, there is a guy at our company. He is good at everything he does but no one wants to talk to him at a company due to his behavior issue. And whenever anyone finds a work which needs talking to him, they drag their feet, make excuses just to avoid him.

Maybe older guys don't fit in the band of recent graduates? Maybe they feel that new guys are reinventing the wheel and dismiss the idea outright?

Just imagine, if someone experienced (or old) (who has years of experience in payment processing) told Stripe founder that their idea is BS because they are tired of payment processing and have seen nothing new.

When an employee is at the best position (age wise), they hop jobs a lot and negotiate higher pay even at the detriment of the company. Yes, it does cost a lot when an employee hops too much. Employers have no option here.

This system will always have winners, you can definitely improve the situation but you'll have to tax the winners, it has never worked well in past.

Polish living standards in the sixteenth century were below those of Italy and The Netherlands but on par with those of England. The author demonstrated that the Polish series began to also lag behind the English values in the seventeenth century due to growth in the latter series. According to a study of GDP made by Wójtowicz and Wójtowicz (2009), people living in Poland enjoyed only around 70 % of the goods and services available to the inhabitants of Western Europe in the sixteenth century. This value declined to around 40 % in the late eighteenth century

source: https://link.springer.com/article/10.1007/s11698-016-0154-5