deepGem

Valid points. Land and expand is a great moat to have. Couple of other interesting points

Their contracts are non-cancellable 3 to 5 year contracts. So retention is almost guaranteed and in banks, beyond 5 years no one bothers to change the status quo. So this is a significant factor when it comes to predicting future revenue potential and profit potential.

On the costs front, hey have to pay a part of the subscription revenue to Salesforce. However, this cost grows linearly in proportion to revenue. The ratio has remained same throughout 3:1 (almost)

Apple can start leaning on their specialized ML cores and accelerators - Can you elaborate a bit more on this ?

I was wondering how this would unfold and looks like things are moving in that direction https://blog.tensorflow.org/2020/04/tensorflow-lite-core-ml-...

If TF models can interoperate with CoreML - boy that'll literally be a home run for Apple cos eventually all ML frameworks will follow suit.

Nope there was no mention of booting another OS. Craig talked about native virtualization that can be used to run Docker containers and other OS runtimes.

I doubt it. If you look at nCino S1, their professional services revenues are not growing as much as their subscription revenues.

PS revenues: 2018 20,094 2019 27,076 ~35% 2020 34,915 ~27%

While subscription revenues almost double every year since 2018. This is also highlighted in their risk factors.

What this indicates is that the deployment of such software still needs significant human touch.

Based on the services revenues and the sales expenses, I am hazarding a guess that nCino is a high touch businesses requiring significant sales professionals.

Which is why, it is key to understand customer retention in such businesses. From what I see, they have a 147% subscription revenue retention rate. That might be a proxy to customer retention but I'll try to get their actual customer retention numbers to be on the safer side.

Overall, yes this is no Facebook or Google. The risk section of S1 is very good and the founders are pretty candid about what the investors are up for.

wow. This makes hanging out on HN worth every minute. I will certainly sink some money into this company once they IPO. Their retention rate is astounding and in a B2B or enterprise sales business this is one of the governing predictors of growth.

It's very rare to find humble low profile founders who just want to build businesses that make money. Ticks off most of my Warren Buffett filters.

My gripe with the touch bar is that for every IDE you have to customise it to see the function keys by default. It's so damn annoying. Why can't the default layout be just the Fn keys ?

The lack of a desire to build is very surprising to me, as an outsider at least. Disregarding politics and everything else, America was (and until now is) a builder's paradise. At least from a taxation and policy viewpoint. I mean, compare US to a place like India which, from a policy standpoint is very anti-builder. Yet, we have seen entrepreneurs springing into action and preparing low cost face shields, low cost masks and what not. Heck they are even making those contactless thermometers.

I am just thinking about this irony, a country of people who have a desire to build, has all the anti-builder policies. A country with the most builder friendly policies has no desire to build. Sigh..

2 kids at home. I am dying to get into my office and work. The interruptions are just insane. I don't think I'll be a lot more productive in the office but I sure as hell will be a lot less stressed. Also my office is very quiet with almost zero interruptions.

Can confirm that Transferwise is the best option out there for cross border commerce. Also stay away from Payoneer if you can. They eat 1-3% of the gross money transferred as commission, depending on the target currency. Also Payoneer's exchange rates aren't the best in the market. This is ridiculous when compared to TW.

The university obsession is not unique to US. It's true in India as well. If you are not from the elite institutions, you won't get past any hiring scanners of top tech companies. You'll still get a job, not the best paying one though.

All books on ACM are free to download till June 30th. If databases is your thing you should certainly check out "Making Databases Work The Pragmatic Wisdom of Michael Stonebraker", creator of Ingres, PostgreSQL and VoltDB and not to mention a Turning award winner.

Yes, many people fail to understand this concept of currency cost averaging. Thanks for the succinct explanation.

Here is my logic as a retail investor. The market has lost close to 20-30% from the peak. Will it regain the peak, for sure. By when ? Hard to predict. All I know is that the rate of climb need not be the same as the rate of the fall.

This virus will bring about significant collateral damage to the global economy. Already airlines, cruises are hurt. Soon, tourism and hospitality industry will be hit. Transport and logistics will be next followed by retail. So the cascading collateral damage is quite unpredictable at this time. Double this with plunging oil prices.

So unless you are willing to wait for 10+ years from now, it's unwise to invest a lump sum amount at once. If it were me, I will just invest small amounts periodically. Like every couple of weeks or so.

Very helpful insights. Thank you.

Yep, but I have only worked on market/users based ideas which invariably need a front end prototype to start with, hence the bias.

Being totally honest here. I don't find any work that I totally can't stop thinking about. There is always something that I need to get done that I don't relish, that I don't enjoy.

Here's how it goes - there's an idea that you think is really nice, you start talking to people then you are hit with reality - no one gives a shit about your idea. Some do, but most of them are like meh. There comes the first punch and you either think they are all dumb or ascribe some positivity to what others are saying and start talking to more people or start working on a prototype. Then you realize you have to build a front end. Welcome to front end hell. You have no fucking clue how CSS has evolved over the last 6 months. The npm package you thought was useful is now outdated or not maintained. So by now you are more or less exhausted. What started as a fabulous idea in the head has now boiled down to 'shit let me get done with this CSS crap and move on'. At this time, I don't even want to think about CSS, let alone HTML. Rinse and repeat this until you find success.

At least, this has been my experience.

It's really hard to be critical of your idea and to be super enthusiastic and passionate about that idea as well. In theory, this works but in practice it's super hard. I have tried it. Either I'll be super exuberant and tend towards over optimism or be super critical and get demotivated like you've mentioned.

I think allocating a specific time to poke holes in an idea might really work.

Another option that I have found helpful is external validation. You talk or demonstrate your ideas to outsiders and they are generally critical of your demonstration. This can be an eye opener some times. You as the idea originator, can then take a real hard look at the critical feedback and test out if that feedback really breaks the idea. There will be a ton of armchair criticism that you can ignore, but you get specific pointed criticism that is super helpful.

Even better is to have a co-founder or someone internal who can take a critical stab at your ideas. Of all the reasons that are out there for having a co-founder this might be the most useful one, especially for experienced folks. It is a bit tricky to make this kind of co-founding relationship blossom though.

My greatest takeaway from fiddling around with machine learning for a past few years is this:

Most of the business use cases are deterministic in nature. It is really hard to shoe in a probabilistic response in such use cases, as illustrated in this article. It's even harder to convert this probabilistic response to a deterministic response (humans in the loop), at a cost comparable to just maintaining the deterministic system.

There are a few business cases that naturally are probabilistic - ad bidding, stock price prediction, recommendations. The giants have taken over this already.

I am still scratching my head to find a use case for a startup to jump in, use machine learning and solve a growing business use case. Probably there are some healthcare related use cases but I have no idea about that domain.

"Training is getting computers to recognise a pattern by giving positive and negative examples"

I think you are only talking about binary classification. Even multi class classification is at a significant state of the art level now. But yes, ultimately this is a pattern recognition process.

I don't think the Corona Virus is as benign as a flu virus. Sure the mortality rate is 1% but that is what we know today. No one knows what this Virus is capable of, in terms of mutation etc, and it is because of these unknowns organisations and public officials are on alert and taking quarantining steps - rightly so.

It is foolhardy to undermine WHO's cause for concern.

I am sorry you are stuck on a ship and I do hope you get to disembark in Singapore.

I still don't understand the pay rhetoric in Canada. I live in Bangalore, and I make more or less the same pay as in Toronto for nearly 1/5th the cost of living.

Why are salaries so low in Canada ? Now at least owing to the US immigration fiasco, more labour should move in and salaries should technically rise - but I don't see that happening. Not sure why.

Competitive salaries on par with California or even NYC will have a dramatic impact on the influx of immigrants.

Man the first edition of this course on coursera had Ullman's lectures. I am not sure if it's the online delivery mechanism or his tonality - I couldn't sit through a single lecture.

One of the toughest courses out there.

One of the most diverse and mutli-cultural cities that I have lived in, is London, followed by New York City. I have lived in Singapore and Toronto as well but they don't match London or NYC.

More than anything I didn't get a tech vibe in Toronto. I was at UofT and the surroundings all the time, yet it felt like the town is mostly devoid of tech. Not the same vibe you get in Seattle or Redmond or SV for that matter.

May be it's my bias but Toronto doesn't feel all that multi cultural. Sure you see people of different nationalities but something feels lacking.

I will add Kettlebell swings to Squat and deadlift. They are one of the best all round exercise for muscle development particularly around the lower back and for cardio as well. In fact, I haven't come across a full body muscle building and cardio exercise that can match Kettlebell swings.

I will frame this and put it on my kids bedroom wall. No one should get addicted to smoking and the youngest minds are the most gullible.

I feel quite the opposite, most of the phone apps will likely to be thick client variants of the web apps, given that the hosts have so much power and storage available. Privacy concerns are gonna force this model where almost every activity having minimal network overhead. We are going back to the desktop app era, is what I feel.

Ok. I can't discern that much of a difference in color or FWIW I don't even know what is DCI-P3. My work is mostly coding and email. So 1080P seems fine. I can't imagine the insane battery life you can get from current batteries, for a 1080p laptop display.

Happy to see that I'm not the only one in this camp. I don't see the need for such Pixel density on the Mac 13inch laptops. What a waste of resources. In fact, am happy with a 1080p matte display.

This is a pragmatic and valid approach. No matter what anyone else says.

Imagine, you told a 10 YO child to do this task. Even the child would ask the same question - how do I know who is a supervisor and who is not.

Not only is face recognition hard, it is almost impossible to accomplish in a factory floor like setting. Not totally impossible but it is really really hard. Face detection is still possible but face recognition is far more computationally expensive. You'll need a shit ton of data and you'll need access to the employee database. You'll need a whole new engineering pipeline to make this happen and of course a team.

Compared to that expense and time, you are way better off getting the company to approve special vests for supes.