Ask HN: What startups do you properly love and are rooting for?

28 points by HakuGulati ↗ HN
I haven’t felt a strong connection to a particular startup since 2012, at least one that I was excited about and anticipating their releases.

What startups do y’all love (can’t be your own)?

48 comments

[ 2.8 ms ] story [ 96.7 ms ] thread
- Mark Cuban's CostPlus [0] to help with crazy drug prices.

[0] https://www.markcubancostplusdrugcompany.com/

The only difference between costplus and walmart/cvs/Kroger/Costco/walgreens generic options is costplus did a viral marketing campaign spamming Reddit, Instagram, tiktok, and news websites.

Costplus is not going to make a dent because they are not willing or able to invest the billions needed for R&D and clinical testing of medicines, and then release the IP/patents to the public. Nor are they going to build and develop the factories needed to make the medicines.

Hence costplus is just an increase in the supply of middlemen, not an increase in supply of medicine. When they develop a new medicine to treat hemophilia and give it away for cheap, then they can be said to have walked the walk rather than just talk the talk.

They manufacture and sell generics with a small markup. What R&D? You just wait for the patents to expire, and the consumer excess is the cost avoided by someone willing to make and sell these drugs with a compressed margin.

https://dallasinnovates.com/mark-cuban-cost-plus-drug-co-top...

> In one example of the savings the startup delivers, the leukemia treatment Imatinib typically retails at $9,657 per month. But through MCCPDC, the price is offered at only $47 per month.

That article reads like an advertisement. Imatinib generic sells for $120 to $150 according to goodrx. $47 sounds like they are eating the cost to gain customers, because if Costco cannot get it at a lower price, I doubt some new business can magically get it at a lower price.

https://www.goodrx.com/imatinib

It will be interesting if they deliver on manufacturing genetics at lower prices, but I do not think they have yet accomplished that. But given Cuban’s history, I would not hold my breath.

Tend to agree. It could very well be a loss leader. They will likely do this for drugs that serve a tiny patient population like this one. They get the goodwill of helping cancer patients save money as well.
I just picked one of the drugs I take.

Finasteride 5mg 30 count at safeway pharmacy: $57

Finasteride 5mg 30 count at safeway pharmacy with GoodRx: $7.32

Finasteride 5mg 30 count at mark cuban's thing: $4.50

When all the other middlemen are basically an overpriced cartel, a new middleman can be very groundbreaking.

Isn’t this why cartel only work with tough enforcement; otherwise, the incentive to cheat is too high
I will believe it is groundbreaking after they prove it from operating for a few years. They might just be temporarily eating the cost for some medicines.

All I know is Costco has a far better reputation than Cost Plus or Cuban, so when someone claims they can get margins lower than Costco, I am skeptical. Cost plus is not even a pharmacy, they outsource all the hard and expensive parts like liability to Truepill. And adding middlemen to get prices lower makes no sense.

https://www.reddit.com/r/pharmacy/comments/wvd964/comment/il...

https://www.reddit.com/r/pharmacy/comments/wfsyif/costplustd...

Now if CostPlus had built factories for medicine and were selling it at a lower price because they developed a novel way to manufacture or have amazing management that figured out how to reduce COGS, I would be interested to learn. But given the nature of Cuban and the fact that they went with a viral marketing stunt with no substance, I am going to guess there is another angle to play here.

Costco doesn’t really have “margins” - everything is more or less expected to cover its cost and MAYBE a bit more to offset loss leaders like the rotisserie chickens. Costco makes all of its profit from memberships.

If they have higher prices than Costplus, that’s not a good sign. They’re a top tier company for logistics and price controls. Sounds like Cuban’s just eating the difference

Do you have any source for your claim about Costco's policy?

I'm seeing that they price items at a very consistent 15% markup in a forbes article.

Costco’s 10-Ks show their profit is almost always equal to about their revenue from membership fees. Which means everything else averages out to being sold at cost.

https://www.investopedia.com/articles/investing/071015/3-rea...

Of course, this does not mean Costco’s pharmacy sales cannot have huge markups, or maybe Costco is paying its employees and suppliers too much, but based on their reputation and my experience over 20 years, I just trust they have the right price to quality ratios for me.

I know I am not getting bottom dollar pricing, but they likely are not gouging and I also know they are a good employer, and thus I believe I will be getting a better product from its pharmacy as well.

With CostPlus, they do not even employ the pharmacists, they outsource to a company called TruePill. How do I know what working conditions they have? Usually when a company starts outsourcing core functions, that means they are cheaping out. And not something I want for my medicine.

If you actually look at their 10-K instead of letting other people tell you what's in it:

2020 Net sales ........................................ $ 163,220

Less merchandise costs ............................. 144,939

Gross margin ..................................... $ 18,281

Gross margin percentage ............................ 11.20 %

Meanwhile:

Membership fees .................................. $ 3,541

https://investor.costco.com/static-files/7ef7bed6-c48f-4687-...

Page 20 shows net income (profit) was $4.002B, which is not far from $3.541B revenue from membership fees. I expect profit to be a little bit higher for 2020 to 2022 years due to volatility in consumer goods prices due to all the stuff going on.

https://www.macrotrends.net/stocks/charts/COST/costco/net-in...

But as you can see, these are increases measured in the tenths of a percentage in profit margin, which would not indicate that Costco leadership is trying to fleece customers any more than they have been in the past.

https://www.macrotrends.net/stocks/charts/COST/costco/net-pr...

Does Vercel still count? It's been a few years, but next to Cloudflare, they are one of the few tech companies really pushing the Web forward, and I'm deeply appreciative of their magic.

I'm really hoping they take over as the premier web host and normalize the use of these services vs raw AWS or GCP. It makes devops soooo much lighter and stabler. It's a night and day difference between this and the in house pipelines I've used.

As a dev, I love that I can just deploy UI and business logic and know that it will work, instead of fighting some esoteric build pipeline every step of the way.

Textio, an augmented writing platform that de-biases language in text, especially for recruitment listings and emails to attract diverse candidates.
Tailscale
I saw a post about tailscale here on HN and recommended them to a friend where he is CTO. The couldn't be happier and became paying customers.
Supabase. It’s an incredible product (greases so many friction points in pg) and the team is awesome. They do such a good job with communicating.
So we know max db size for Supabase?
I don't know off the top of my head, have you tried emailing them?
Disclaimer: I work at Supabase.

Free projects: 0.5GB.

Pro tier: 8 GB included, then $0.125 per GB.

I can say that we helped a customer migrate a DB of over 1 TB over from RDS to Supabase. Feel free to contact us if you have any questions.

Is there anything approximating a hard limit at the moment (AWS Aurora is at 128TB at the moment). Do you do anything with Citus or CockroachDB or normal partitioning behind the scenes?
Aether. They make electric 2 wheelers for the Indian market.
Framework laptops! For reasons that are hopefully obvious to this crowd.
Purism. You can debate all you want about some of their more than questionable business decisions.

That being said: they’re making phones that run an unmodified mainline Linux kernel.

> more than questionable business decisions.

What are some of those?

One thing that comes to mind is coming out with a bunch of other products before backers had the first shipment of the laptop they made. Curious if there are others too.

Fulfillment of Librem 5 orders has notoriously been behind schedule – which is unfortunate, but shit happens. Of course, some customers didn’t want to put up with a years-long delay. So they opted for a refund.

At one point, Purism changed their policy. Under the new policy, customers wouldn’t be entitled to a refund until their order is actually ready to ship, even if that happens several years down the road. For example, a customer won’t be able to claim a refund for their 2019 order until e.g. late 2022/early 2023, when their order is ready.

The questionable thing is: Purism has retroactively applied that new policy to earlier orders, too. So by Purism’s logic, customers whose order is years late can’t get a refund immediately, even though at the time of the order, the refund policy did grant that right.

People are constantly being furious about that, and I wonder whether Purism is doing themselves a favor with this, or if their behavior is even legal.

I’m rooting for Oxide - big fan of the past work of the headline crew involved, and their interesting and approach to on prem hardware.
I really love LaunchDarkly as a company. The feature set is solid and has served us really well, and from the outside, the culture seems fantastic. Their sales and partner success teams were much more responsive than competitors. We're very happy with them, and I continue to root for them as much as I can.
Repl.it- they've been moving at a breakneck speed recently and have grown a ton from what used to be only useful as a REPL.
comma.ai. full manual driving is barbaric and unsafe.
Descript for podcast editing. I used it for an internal podcast last year and got to experience childlike glee once more.

It was extra special because I work in nlp but so many nlp applications just don't excite me

OpenAI - It is doing wonders, especially with Github Copilot. Netlify - It has improved my productivity. Reddit - It is the face of the internet and I want it to continuously improve. Nvim - It's a technology and community upstart.
Tenstorrent - Jim Keller.
Rooting for Manara(YC W21)[0] to the point people I talk with ask me if I'm a founder or have a stake, and talk I do.

https://manara.tech

Unless you live in North-Africa or the Middle East or places like that, you can't do justice to what Iliana and Laila are doing. Many of these countries are prisons with flags. Isolated from the world with pretty much useless passports. Isolated from monetary flows. Isolated from the global job market. Scrutinized at every airport. These people's only sin was to be born there, and they're paying a dear price, especially with the cultural values of sticking by parents/grand parents. Some of these countries actually forbid their citizens to build an estate or own something in another country, many have non-convertible currencies that make it hard to do business and black markets, even black markets for visas.

In other words, if you were born there, you're screwed unless you muster a disproportionate amount of grit to get to a baseline.

Go Iliana, go Laila!

If you know people in North Africa or the Middle East, that's a great way to improve the situation.

[0]: https://news.ycombinator.com/item?id=25849054

Thanks so much for the shout out! We believe that the concept of "undiscovered talent" should be eliminated globally - and that this region in particular has been overlooked for too long. In Palestine for example, 52% of computer scientists are women, but 83% of them are unemployed.

Running a startup with a social impact mission is a dream come true. I hope more people will launch businesses aiming to solve important challenges. Our world needs the creativity, talent, hustle, and funding that are so easy to find in startup communities.

Fly.io, I really respect them and their awesome "light vms". Their blogs are fantastic and one of these days I'll have a reason to try them.
Rogobi - https://www.rogobi.com/

Scientists have a big problem. Abstract search can take up a big piece of any researcher's time. They often farm it out.

The quality of results can vary depending on who does it. Researchers in sibling specialties often use different phrases for the same concepts.

Rogobi manages cross-disciplinary search by collecting terms across disciplines and guiding your search as you go.

Like other abstract search engines it also manages PRISMA details, collects and generates result reports for your clients, and coordinates multiple researchers sifting the same results.

The startup's difficulty is finding the sweet spot of a customer with money that also recognizes the pain point (shallow/silo'd searches due to search term fracturing across disciplines).

OpenAI. I love how fast the world has gone from "lol AI could never do that" to "AI has unfair advantages!" Much of it is based around OpenAI demos like GPT-3 and DALL-E 2.

They've set the benchmark and expectations, and put it ethically high, even if it means ironically not being as open as their competitors. Just like Google set advertising and free software as the norm of the Internet, OpenAI set safety as the norm for AI. If many other companies held the lead by now, the internet would be full of sex chatbots, fake reviews, deepfakes.

When you start seriously playing with DALL-E you realise how pride and limiting they are.

I wanted to create illustration for books, produce high end book collections and redirect part of the profit to a charity in Benin and rural China.

It’s hardly possible, you can’t generate any forms of violence. Yet violence is part of life, and pretty much half of literature.

How are supposed to illustrate Lolita or do justice to Dostoevsky with these constraints?

I seriously hope a competitor take the lead.

supabase, typesense, zerotier, tailscale