Here are 6 startups for you to read about and engage with where interested. The initial order is random.
Lernit (YC S21) - Corporate training program for Latin America - https://news.ycombinator.com/item?id=28049505
StandardCode (YC S21) - APIs to easily comply with child privacy laws - https://news.ycombinator.com/item?id=28049504
Scispot (YC S21) - Workflow automation for life science - https://news.ycombinator.com/item?id=28049501
Muse (YC S21) - Allow anyone to build 3D websites - https://news.ycombinator.com/item?id=28049502
Ruth Health (YC S21) - Digital, at-home post-pregnancy care - https://news.ycombinator.com/item?id=28049503
Deskimo (YC S21) - Book workspace by the minute in Singapore and Hong Kong - https://news.ycombinator.com/item?id=28049507
Raphael and I have been working remotely for most of the past 10 years. During the pandemic, all of our friends suddenly saw themselves thrown into a similar position and we've observed how they adapted. Many don't have a suitable home-office setup and while for most the pros of remote/hybrid work clearly outweigh the cons, having access to near-home desk space would solve the downsides of working from home for most.
We give users access to a wide range of professionally run workspaces in central business districts as well as in residential neighbourhoods. By giving users access to near-home workspaces, we give them an opportunity to work for a few hours and not waste precious time on commuting. We also have locations close to schools, shopping centers, and other places where our users need to go throughout the week.
There is no need for a membership, upfront payments, or long-term commitments. Access works similar to ride hailing: location-based and instant. A user scans the QR code of the app at the reception to start the session and scans the code again to check out. We charge at checkout for the duration of the work session; the maximum charge per day is capped at the rate of a day pass at the specific location. Our customers consist of companies who provide Deskimo accounts for employees, and individual users who prefer our flexibility over the cost and commitment of a coworking membership.
Do you plan to build/rent more traditional/conventional offices?
I personally cannot imagine a room packed with professionals at different companies on lounge-chairs and sitting around tiny tables. However, I might not be have a good understanding of Singapore/HK culture.
Eskimo isn't an offensive word, and is only considered inappropriate when its used to refer to Inuit peoples. Which this isn't.
: at this time.
The name is super witty and probably nobody in SG/HK cares, but I don’t think the people generally called Eskimos like being called that.
Then again, Volkswagen doesn’t get flak for selling Tuaregs as far as I can tell.
Alex and I started building 3D websites because we were bored of building the same old static websites that we have been contracted to do for the past 6 years since high school. Building websites sucked and was boring. Fortunately, Alex discovered the power of three.js. Between us, Alex was the first one to start building websites in 3D and frankly, I was jealous because I saw that the code for 3D websites was far too complicated for me to grasp quickly and I felt left behind. This is why Alex and I started Muse, to make the 3D internet accessible to everyone and easy to use so we don't have to keep building static websites.
The way we solved this problem was by first building an open-source framework called SpacesVR that made it very easy for React developers to start building 3D websites. Several months later, we built a no-code editor that works similar to Squarespace and Wix built around this framework. Using our framework, our team has built over 200 3D websites by hand. Since launching our no-code editor, we have seen over 40 websites published. Also, a crazy thing happened the other day when Alex was scrolling through Twitter. He stumbled upon a 3D website that used very similar control mechanisms to our websites. With a little bit of investigating, we found out that a virtual gallery project, 3XR, started using our framework to build virtual NFT galleries for all Mintbase stores!
It has been exciting to see our no-code editor and open-source framework grow. I encourage everyone here to build a 3D website, try it out, you may enjoy it! You can check out our open-source framework at https://www.npmjs.com/package/spacesvr or go to https://muse.place/to build a 3D website with no code. Have fun building!
Generally an interesting field to work in, lots of uncharted territory in terms of UI best practices and user expectations. Good luck!
Not quite sure what sort of site, on which I was attempting to actually accomplish something, I'd be happy to find something like this on, but there's probably some use I'm not thinking of.
You can include a link to a demo and the showcase sites elsewhere.
It also feels what you're trying to sell is full 3D website templates, when in reality it would probably be more viable as an embed or a section of a bigger site.
An example of a successful site that is built on similar tech is decentraland
Surely there is some UX technique you guys can come up with to free the mouse and improve this.
Question for you... what is the best 3D website on the Internet? How do they solve this?
All the best to you guys!
Also, one of our favorite websites have been https://nurtu.re/
It is built by Active Theory for Porter Robinson. I believe this could be the future. I guess they get away with it because they are in third person.
This is mostly just a side project now, after 6 years of work. We have our own 3D renderer, but are using our custom Cesium here.
To be scrupulously fair, the fact that a site can do this, even on purpose, is both a browser bug and a operating system bug (denial-of-service security vulnerability in both).
Have you considered how you intend to provide accessibility on top of this model of interaction? I have some thoughts for possibilities involving HRTF audio for blind users, but the deafblind experience is going to be ... tricky.
Of course you'll need to do similar work to expose the semantics of one of these experiences to search engines, so for once I am not just begging for accessibility merely for us poor blind folk.
I'm curious as to how you internally represent the scenes, and what sorts of research you've done into precursor technologies such as VRML.
If you have questions or would like some ideas of where to get started with accessibility, feel free to reach out! My email is in my profile.
I am building a Reddit for China. I can't help thinking about builidng a 3D version.
This is like, what can I say, muse!
If you visit at least once without tracking protection, and then turn it back on, and visit (even with cleaning all caches), then the site loads correctly.
I've seen some comments about the performance, and I just wanted to comment that I just tried the site out on an old Chromebook Acer R11 (Celeron N3150) and while the perf on the opening environment is a little bit hitchy, the Balloonski room for example ran perfectly well. Little machines are capable of a whole heck of a lot these days! Even in the browser!
There is a steep hill to climb before a 3D web really lands in a way that fosters an ecosystem of content creators. I suspect that it will take a different breed of web browser to truly realize an XR web, one with built-in, high-level APIs for composing 3D content. Ultimately, you need many of the built-in features of the 2D web, and approximating those things via a canvas is tantamount to rebuilding a fair chunk of the browser.
My Calendly subscription ran out :)
Here is our updated link: https://calendly.com/ben-792/userinterview
- Would creating a 3D website hurt my SEO?
- Would the content be somehow accessible for users who are entering the website on devices that don't support WebGL? Is there any fallback?
- A 3D website looks cool and creative but does it really engage users more than a standard one? Would it increase my conversion rates or keep people longer on the site (subtracting the time to figure out the controls)?
- Most of the end visitors, those that visit our creators websites, are actually mobile users and they have an average session duration of 50+ seconds and so far we have very high retention with our creators so I think there hasn't been to big of an issue with accessibility. I tested our website on a iPhone SE and it ran great on good wifi. All modern browsers support WebGL. Our fallback would probably be around bad wifi.
- We only use 3D web sites to run our funnels. I always tell people content is king at the end of the day so I guess you will have to just try it out and see how it goes. Right now, it is hard for us to share with you robust analytics via conversion rates and CTAs so I recommend using a service like BITLY and attaching bitly to the buttons so you can track clicks. Here is one of our funnels: https://www.muse.place/pandassaurus
The funnel was built completely with builder tools, so you have the same tools I do. So at the very least I can tell you that our growth/marketing strategy at Muse is 100% reliant on our own product if that helps.
I really like these fun alternatives to ordinary web sites even though the UX designer in me is having a heart attack because of the countless usability and accessability issues
Reminds me of VRML.
I am working on building a company website right now with bg info which should come out later this week. Being built using builder tools!!!
We deliver 30-minute, online physical therapy video sessions to help the 83% of postpartum womxn with moderate-to-severe Pelvic Floor Prolapse—in other words, who pee in their pants after pregnancy. Birthing people deserve a life without postpartum pain and incontinence.
With our personalized, 1:1 telehealth Pelvic Training + Recovery Sessions, moms worldwide can get back to work, sex, and normal life 5x faster—all from home.
From SF to NYC to Singapore, patients simply share health information, get scheduled, and then log on for our video sessions during this $10-145/session sliding-scale Pelvic Pilot program. By taking our Training + Recovery Sessions 30 min/week from home for 2-6 months, womxn see an increase in strength, reduction in pain, and greater bladder control—versus spending often double the price and 3+ hours door-to-door on each Pelvic Floor Physical Therapy appointment in-person. Onboard here to get scheduled (https://bit.ly/RuthHealthOnboarding) for your first session!
The 5x reduction is as opposed to postpartum recovery without any form of pelvic therapy or training as intervention.
Her primary OB does not offer Pelvic Floor Physical Therapy or anything like our Pelvic Training + Recovery. These are most often found at a PT or with a fitness trainer. Our methodology is proprietary and speeds up recovery with a targeted approach and 1:1 sessions—as opposed to videos, group sessions, or longer PT appointments that have low adherence due to a long time commitment.
We are for birthing people with acute issues! But we are also a great preventative option for those who never want to get there in the first place. Pelvic floor prolapse gets worse over time and with the birth of more babies by an individual.
We hope to make maternal healthcare better for every body—including your wife's!
Congrats on expecting :] and we'd love to help. You or your wife are welcome to sign up here and we'll reach out to schedule! https://bit.ly/RuthHealthOnboarding
COPPA regulation makes it difficult for companies to cater to children under 13 without building in safeguards to collect and manage parental consent and to block third-party sharing of data without transparency. As a result most companies just prohibit users under 13 from signing up for their services, either as a policy or by having an age gate that prevents an under-13 child from signing up.
However, with a record number of under-13 users on the internet due to the pandemic, behavior is permanently shifting and many companies now want to cater to all audiences. By providing APIs to manage the collection of consent, we allow companies to safely acquire underage users while minimizing friction in the sign-up flow.
Our customer-facing REST API presents a few high-level resources that are needed for collecting parental consent and ID verification for children. Customers typically make 1 API call to create a profile for a child, and then 2–3 more calls to collect parental consent or verify age and ID if necessary. Afterwards, we push data to a webhook endpoint to notify the customer when consent has been collected or an ID has been verified.
Our business model is based on usage — we charge a small fee per user for whom we need to collect parental consent or verify age and ID.
We'd love to hear from you, especially if you have experience with products catering to children!
You mention COPPA and GDPR but which country regulation do you cover (and in which language?). How do you ensure that the Parents giving consent are the parent of the child (and that you are not getting spoof?). How do you keep up with the changing regulations, with some country having multiple layers (Union, Federal, State, City, ...).
I like the focus on Childs, but you could also extend it to do KYC for financial institutions, the market may need a cheap and reliable solution.
re: spoofing. this is a tough problem to solve and no existing solution guarantees this. we would be able to see suspicious behavior though like multiple verification attempts, time between submission and verification, etc.
Changing regulations is tough. We work with a privacy lawyer to keep us abreast of changing policy and have researchers maintain a database of applicable state, federal, and international laws as well as recent lawsuits.
Financial KYC is interesting for sure. Right now we’re focused on child privacy and very much believe there’s a lot of work to be done there still.
We certainly do work very closely with each company to work through their product and integrate as seamlessly as possible. whereas we don’t offer an official security review as of now, it is something we’re considering for the future. would love to chat, DMing you now!
According to studies, 74% of CEOs worldwide are concerned that a lack of essential skills in their employees is threatening the future of their organization. With the increase of remote work comes the challenge of keeping teams aligned and pushing in the same direction.
When I graduated from college, the biggest learning platforms were Coursera and Udemy, and it hit me that there is an immense gap between what you learn in an on-demand course and what you do at your job. Usually, these courses were outdated and had no relevance to what employers wanted their employees to learn. So I saw a huge opportunity to create Lernit, where companies could ensure that their employees were learning valuable skills that could be used at their job.
We allow companies to develop online courses on the specific skills their employees require for accomplishing organizational goals and personal growth while staying focused on priorities. These priorities are easily visualized and aligned to the company's main objectives through OKRs (Objectives and Key Results for a specific period). Each individual can check in and track progress while receiving feedback and mentoring in a collaborative environment.
We offer an all-in-one solution for HR teams where companies can set, align and track goals, evaluate performance, train and develop their workforce and build culture without the trouble of managing different tools at expensive prices.
We are eager to help companies reach growth and more than happy to answer any questions you may have about our talent platform, so if you have any or are just curious please leave a comment!
What are you going to do differently? How are you going to gamify this in a way others don't?
I've worked in lots of places and interacted with many software portals that try to do this right. A lot of them falter by not keeping up with the actual current state of information within an organization (especially a rapidly growing one). Many of the courses end up outdated and they're really only useful for onboarding.
It's probably because of the time it takes and the friction needed to create a course or update a course. Because of that friction and because it's easier to just post to a Wiki / Slack channel / internal blog / etc. the small updates and changes, these courses sometimes languish.
Just questions that come to mind. And congrats on the launch. Wish the best for you!
Currently there’s no solution in LATAM for medium sized companies, and with more companies working remotely the need for tools that will help keep teams aligned and focused on priorities while filling learning gaps is growing as well.
The friction needed to create courses for specific skills or roles is the whole reason why Lernit is now a reality, we want to make upskilling available for every role and level, and thanks to our authoring tool that enables companies to create courses in minutes, our marketplace with up to 6 new courses per week and our strategic alliances with great universities that gives our users access to learning programs is that we are able to keep up with a fast evolving market and organizational needs.
Please let us know if you have any additional feedback or questions, we will be happy to hear back from you.
My only piece of "unsolicited advice" would be: raise prices and focus on value added. Anything that is value added will be difficult to copy and will be your difference with anybody else: community, cohort building, etc.
Best of luck!
Life science companies often have no choice but to rely on a laundry list of tools to stay up and running. Electronic lab notebooks, LIMS (Lab Information Management Systems), Google Docs, Asana, Notion, and Airtable are just some of the tools that are used out of necessity. As a result, most of the biotherapeutics, diagnostics companies, contract research organizations, and labs struggle to stitch together disparate data. They end up scattering their sample, inventory, project, and protocol data in various systems. The disparate data adversely impacts the experiment's failure rate. Not being able to connect data also impacts the data integrity and regulatory compliance for fast-growing bio companies.
We let you personalize your workflows using an orchestration layer. A lab can track and manage its samples and inventory, plan and prepare experiments at the desk, and execute experiments at the bench using Scispot. We have project management features to easily visualize research and operational activities in a calendar and a kanban view. We also have APIs for developers and data engineers to plug in existing systems. For instance, customers can integrate their batch runs and inventory to their legacy systems using Scispot’s API. Scispot maintains a thorough audit trail (eligible for CFR part 11 compliance).
Customers are using Scispot to connect their inventory and chemical library with experiment execution. Customers also use Scispot’s template library to create protocols in bulk and connect the protocols with the inventory.
Scispot is founded by three founders, Guru (a molecular biologist and a life science tech veteran), Nash (automation expert), and me (the product guy who loves building workflows). We’d love to speak to any of you that are curious about what we’re doing or if you have any ideas/challenges for us!
HR type functionality, Employee management and financial management are operations that are currently well covered by many of the ERP solutions out there in the market. We don't immediately have any plans to build these features right away, however, we never say never ;)
If a lot of our customers would deem these feature sets necessary, we might devote our engineering efforts towards either building them ourselves or integrating with a major ERP solutions providers down the road
2) What are the things you offer that nobody else does?
3) How long have you been developing the product?
4) If it is only three people, why would a lab choose you over established players?
2) What are the things you offer that nobody else does? There is no other no-code workflow automation in life science space yet.
3) How long have you been developing the product? One year
4) If it is only three people, why would a lab choose you over established players? We only mentioned our founding team as part of the post. We have a team of highly talented engineers, scientists and advisors without whom Scispot wouldn’t be the product it is today. Customers choose us because we are more agile compared to bigger companies. Our agility comes from the orchestration layer that you can use to design your own life science workflows. For instance, every customer has its own different flavour of Scispot with different workflows, and metadata.
Happy to set-up a call to continue your discussion.
Maybe something along the lines of "Launch YC x 6: 3D Web; Training; etc.." could make it more clear how many launches are in a thread?
Possible to group by theme or something?
Or just leave out the specifics. The 2-word pitch is cutting too much to be any sort of useful.
Since examples are the best way to learn, I would be interested if anyone could come up with a better title for this thread. It's not obvious how to do this!
I think on balance you're probably right that excluding a startup from the title is a problem though -- so yeah, in hindsight, I think you made the right choice.
Another thought I had was "Meet S21". Better? It would free up a char! "Meet YC S21" is too long.
"YC S21" would be the shortest. That is, something like "YC S21: Foo, Bar, Baz" with a word or phrase for each startup. But I'm not sure if that would convey that it's a launch thread.
Another possibility: "YC Launch" instead of "Launch YC".
Our original idea was "Meet the Batch" but that's way too long.
Launch YC S21 - Batch thread #3
What a strange problem to have. Why not just increase the limit? Why can't you have a special title with a higher limit?
>Actually, we pretty much don't change that ever.
A classic mistake.
How about mirroring the Who’s hiring format?
>Launch YC: S21 (August 3)