![YouTube video for the AMA YouTube video for the AMA](/static/eca94b63a6c098442a701881c32e1051/c1d72/ama-screenshot.png)
I was just saying that it's nice to be here, chatting with you guys on video for a change. We're dialing in from Ente's office in Bangalore. It's 9:30 PM here, which is why it's a bit dark outside. I've got Laurens here with me.
Hi, guys. My name is Laurens. My path with Ente started, like, almost 3 years ago as as a customer . Same as you guys basically. I was in the community and that's how the ball started rolling. I joined as an intern and now I'm working full time here. Also moved to Bangalore, and I work on the the ML stuff. Right now, faces, is the main thing that I'm working on.
Let me quickly give an introduction to what we are planning to do today.
You guys all submitted your questions. Thanks a lot for that. And you also voted on it. So that's great. We need some kind of ordering on how we go through the questions. I'll be asking the questions. I have them here. Vishnu will be answering them.
Two introverts sitting in front of camera. So, it'll be like a casual easy conversation. But we just really hope that you guys like it. We really enjoy how much is happening in the community. And that we are thinking about how can we engage with the community more. And we thought this would be a nice way to do it. So, yeah. Let's see how it goes. If you guys like this and find this useful then I think we can do this more regularly, like maybe once a quarter or something. We'll figure it out. But, yeah, for now, let's get started.
Like I said, everything is in order except for the first question just because this is a nice question to kind of start off with.
So the first question is...
What's the story behind Ente? Why are you building Ente? How was the company founded?
That's a nice question to ease into the AMA.
For those of you who don't know, before building Ente, I spent a couple of
months working at Google in Zurich, Switzerland. Not the best time of my life to
be honest. Because not a lot of things happen at Google. It's a very slow moving
company. Before Google I had worked at a bunch of startups. Really nice places
to work at. In general, I'm very high energy person, and I like being
productive. So for someone like me, it was not a mentally great place to be in.
And, the other problem with Google is that they don't let you own the pet
projects that you work on. If you want to work on something, by default they own
the IP. I'm not sure if that has changed now. At least back then there was this
elaborate process of getting approval if you wanted to own whatever it is that
you are working on, and I had a lot of spare time at hand. If there are any
Googlers here, you can check out go/iarc
, you will know what I'm talking
about.
What would happen is, this committee would go through your proposal and would approve it if it turned out that, whatever you're trying to build did not compare to any of Google's business units. And the problem with Google is that they're practically building everything under the sun. And it makes it incredibly difficult for you to , get through the process. So it was a very frustrating time for me, not being able to build anything. So I knew that I had to quit and do something else. And the question was what.
A couple of things were happening at that point. Me and my wife had just gotten married. We had just moved to Switzerland, and we were traveling around quite a bit in Europe and clicking a lot of photos. And, this is around the same time when I realized that I should probably not give Google access to all of my photos.
So I did the next best thing. I bought an iPhone, moved all of my data to iCloud. Surprise surprise, my wife could no longer access all of my library. Because she was still on an Android device. Apple being Apple has a very tight lock on their ecosystem and where their data can be accessed from.
This completely broke all the sharing workflows we had for us. So we had to get out of Apple.
The next place we went to was Dropbox because that is supposed to be the cross platform guy. They do have apps everywhere. But the problem with Dropbox was that they don't care about photos and much less about consumers. So it's not a great UX.
Which is when I found out that there's nobody else solving this problem of building a safe space for normal consumers like me to preserve my photos.
That got a lot of things rolling in my head.
For me the other aspect of this was that I'm very sentimentally attached to my photos. I grew up around a lot of photos. My dad used to run around us as kids with his Kodak and Minolta cameras. So for me, this is like a very personal thing to build and that checked a lot of boxes because I realized that this is something that I would not give up on, because of the sentimental angle to it.
For me, that is a core value as well. To play really long term games.
In 2020 I came back home to Kerala, India. This was right when COVID hit.
I started working on Ente.
The first project that I started working on was quite different from where we have ended up. It was a hardware device. It was a private Google Photos server.
If you guys want to check it out, the landing page for that is still up. Yeah. You guys can check out at https://orma.in
It was just literally like a private Google photo server.
And then I realized that the problem with building hardware is that it's incredibly hard to monetize and scale. Especially without VC funding. And, venture capital was not something I was, very eager to explore at that moment. A year into building I realized this is not really gonna work. Then I pivoted and I started building a cloud based service instead with a layer of encryption between because I didn't want access to any of your photos. That was the whole point behind this.
We launched the product a couple of months later on Reddit first and then HackerNews. Somewhere in between, Bob, Bob is Neeraj, our cofounder, and CTO. Bob joined, and that is, the best thing that happened to Ente and to all of us. That gave a lot of momentum. And since then, it's been 4 years. We've just been building, and it's we've had a good time. That's it, so here I am telling you guys about this story.
![Photo from AMA Photo from AMA](/static/f8a1a805520e86931ca02a58a006f8db/d6032/ama-1.jpg)
Will Ente always treat GrapheneOS users with the same importance as other users, like you do now?
This is also a really nice question because it, gives me more of an excuse to tell you more about my origin story.
So, see, I'm a programmer. I have been building stuff since high school. I think 8th grade is when I started building websites and templates and just building, like, good looking stuff and which is why I'm very very finicky about pixels in general. I think Pragadees will know that it makes me a very difficult person to work with when it comes to design.
Then I went to college in 2009, which is Android started gaining popularity. This is when Android phones started hitting, India as a market. And, I found that whole thing really fascinating. There was this programmable device that you could handle the palm of your hands. This was very different from what we were used to. Symbian is what we had until then. Android was on a league of its own.
I emotionally blackmailed my Dad into buying an Android phone at the point. It was a Samsung Galaxy S3. It's a mid-end device. It was $150 at that point. $150 was a lot of money for our family at that point. We come from a very middle class household. And I'm very grateful to my dad for taking such a blind bet on me in general. So dad if you're listening, thank you!
That was my foray into open source as well, because Samsung has stopped shipping OTA updates for my device.
So the first thing I did was compile the latest AOSP, for my device, and that's how everything started. Then I joined this froum called XDA-Developers which is where Android developers hung out at that point. That's where where CyanogenMod was born. LineageOS is a fork of CyanogenMod.
That's where stack envy kicked in. I felt OS was not cool enough, so started working on kernels. I used to maintain a bunch of kernels for Sony Xperia devices. This is how I also got my first job.
Long story short, I look at all of these projects with a lot of fondness because this is where I grew up. Be it Lineage, Graphene, Calyx, it's really nice to see these communities shape up and push these projects forward and not let AOSP exist in a silo.
So for those sentimental reasons, I value GrapheneOS quite a bit. I think what they're doing in general with pushing technology forward, pushing privacy and security forward. It's really nice to see. It takes a lot of balls to make the promises they make, and deliver on those promises. In general I look up to them. This is the sentimental angle to this.
From a business perspective as well, from support tickets there are quite a few users on Graphene, who help us out with debug logs and just in general a very helpful community. So we get money, we gotta keep supporting the community.
Long answer short, yes. We really appreciate Graphene and everyone that is helping us give in the community. Especially the Graphene people are super kind, super helpful. Long story short, we will keep focusing on Graphene.
Will the semantic search and face recognition get faster and better over time?
Yes. We're a company who's very heavily betting on ML running on device. We believe computing will get better and models smaller. And there are not a lot of companies who are getting these models to run on mobile. The reason we're doing it is because we believe this is the future, and this is where we should be placing the bets. So it's a very strong yes.
Will Ente make a Contacts app?
v1 of Photos first. Because there's a long way to go to get there. We don't want to try and do too many things.
We do feel the need by the way. We personally feel the need. Bob has been pitching this forever. It does make a lot of sense.
Shipping the first version of any product is the easiest thing. What happens after that is the harder part. Providing support, iteratively improving the product. Lot of work that goes behind it.
If we were to stick Ente's brand on any product, we would want that to be something which is very good. Like, it gives you an excellent customer experience. And, we are not at a point where we can do that . But we definitely want to do it.
Even with respect to photos and contacts, there's a lot of synergy, like, in terms of how we want to share photos in the future. Like, if you could set up a system like Signal where it's one tap sharing, there's a lot of workflows it would enable as well. So, it's a yes, but, not right now. Soon.
![Photo from AMA Photo from AMA](/static/06ff3f341b3eb630e7765a7717f2fcd8/d6032/ama-5.jpg)
Improving the video player
I feel you. The level of support we have for videos is not great. There's a lot of work to be done, especially on the streaming side. It's massive amount of work. So we know exactly what needs to be done, but it will take us a couple of months to execute.
So this is something that we've parked for, until we've done face recognition. We will do it because videos are a first class citizen for us. When we get to it is when we will solve the problems you've mentioned. Those are easier problems to solve, with playback speed and zoom. It's just client side stuff that we have to do.
So will do! Soon!
Custom lockscreen for Trash
Custom lock screen, yes. We're already working on it. Trash, not right now. We'll have to figure it out. Thanks for the feedback will keep this in mind.
Gender recognition within face recognition
I guess this is like the one question that I'll answer since this is basically is what I am working on. First of all, like the disclaimer, this is kind of basically one of the reasons that it is still in beta. We are still working on it. And I have seen this before also with me. I think usually it is family members though. And sometimes it's kinda interesting to see like, oh, I guess I look like my mom or something. But yeah, I understand also that maybe this is not what you want. You want just direct correct suggestions. And for that I can say, yeah, we are working working to fix that and to get that better. Even now that it's in beta, we are, experimenting with a different type of clustering, which we expect will already solve this. So that's kind of what I am aiming for now. Have the better clustering and that should ideally solve this problem.
I don't want to get too technical into it. But basically, we are working on it. Like, and even if the clustering won't work, there's also other ways of of fixing this. But ideally, the clustering itself should fix it. Yeah. I don't want to get too technical. But, definitely, we are working on it. Definitely, you can expect suggestions and faces in general, like we said before, over time, it will keep going getting better.
Offering a cheaper service
They mentioned about us, promising cheaper plans. I don't remember promising cheaper plans. But I do think we've to make Ente more affordable for a lot of folks because there is a barrier to entry. And, it's not just because Ente is end to end encrypted. It's more because we keep 3 copies of data. We have 3 replicas.
Google can definitely subsidize this a lot more. I don't know why Google is charging. I think they're charging just to not seem suspicious.
While we've got a sustainable business to run. But what we would like to do is, there are multiple ways to approach this. One is reduce number of replicas or offer a plan with less replicas. The other way is to experiment with pricing pricing and maybe offer smaller tiers, like, cheaper tiers upfront.
If we play with pricing I think a lot is possible. We've not been experimenting as much as we should. We'll be doing it next quarter. So let's see. The idea is to try and make Ente more affordable. That's definitely the goal.
Which platform is the hardest to develop software for?
It's hard everywhere.
Relatively perhaps server is a bit simpler because, because you just pass around encrypted blocks. So it does make things a little simpler.
But, there is other complexity. In terms of maintaining infrastructure, maintaining reliability, resilience. For a company of this size or for a product of this maturity, we have incredibly mature server. I'm not sure how many of you guys are looking to our source code, for museum - museum is our server and a lot of it doesn't show. The maturity of our infra is really high. So it's it's not like that is very easy either.
But if you were to compare it with, say, mobile, right, wherein compute is scarce resource. Mobile is much harder to build not. And even with web, a lot of the APIs that we want to use are cutting edge and do not work in half the browsers.
Client side engineering in an end-to-end encrypted landscape is very difficult. It's hard.
Also because of the expectations you come with - the benchmarks that you have for Ente is apps like Google Photos or Instagram. To meet those benchmarks is not easy.
But hard games are worth playing. There's joy in doing hard things.
How big is the Ente team?
There are 12 folks on the payroll in one way or another.
There are 8 of us who are working full time on Ente. 4 folks who are working on Engineering.
There is Ashil who works on mobile. There is Laurens runs all the ML stuff. There is Manav who looks at web, desktop, infra. There is Bob who practically does everything. Then there are other 4 of us who look at product and support and marketing.
Then there are 4 interns, or folks who are working part time. Prateek is the guy who has been maintaining Auth for a couple of months now. We've got James, most of you know James. He has been working on a lot of cool projects. Like cast and passkeys. He is working on something cool right now, I can't wait to show you guys.
We've more folks. Aman was the one who shipped share-link-preview feature. We've got Atyab working on dashboards.
There are a lot of volunteers as well. Recently Sooraj has started contributing to our blog and helping us with content in general. There is Fr_g (Brogio) who has been helping us run Discord since god knows when. Lot of folks who help with translations.
There are a lot of really nice folks whose kindness we are running on. Friends, like Rahul who has drawn the ducky, and continues to draw illustrations for us.
Coming back, the core team is 8. And it's intentionally small because it's much easier for us to move the way we want to move. We feel this is effective, and we hope you feel this is effective as well!
Vision for Ente 10 years from now
That's a loaded question.
10 years is a very long time frame. A lot of things we would like to accomplish. Primarily we'd like to be the photos company, and that's a very aggressive mission. I feel it takes a very long time to build very solid enduring companies, but I would like to build that company.
On the company front, we would also like to be the company that families trust to share and pass down data.
From a tech perspective, we would want to be the guys who are pushing local AI forward.
From an internal people perspective, I would want to work on making myself replaceable in this decade so that even if I'm not around things just keep running.
From a product perspective, we want to be that very friendly consumer brand that people attach positive emotions to.
This is a very loaded question, and I could keep going on, but I hope you get the gist of where we want to be.
What privacy service would you like to see developed?
Contacts, like we discussed, we defintely want to see that.
Personally for me, I take a lot of notes. I journal quite a bit. I write down quite a bit of things.
For me, when I used to use iOS, I used to use this app called Craft, which used to work pretty well. And it made writing very delightful, and I used to enjoy writing. Since I moved to Android, I've been using Notesnook and Standard Notes on and off. And both good products, but, there is a level of fluidity or magic that Craft presented which I miss quite a bit. And, I think, at least at Notesnook Abdullah is shipping crazy fast. So I think we might get there.
In general notes, from a personal perspective. I would like a very very fluid notes app that offers end-to-end encryption.
But we are not building that, please don't get any ideas!
How many users does Ente have?
We have around 50,000 users.
I would like to be a lot more open about our metrics. I would like to start behaving like a public company in a sense. Where in we publish quarterly reports, earnings, and metrics in general.
I would love to have a dashboard like ente.io/public
. It will happen, soon. We
just haven't gotten around to it. Not because, we have fiduciary duty or
anything, but just from a perspective of transparency. I think that's what
everyone deserves. We should be open and we will get to it.
If at all this question was a proxy for whether or not we are gonna survive or whether we need extra investor or VC funding to make it, the answer is no. We are good right now.
In fact the last month is when I took my first salary in 4 years. But all of us could be earning a lot more. We should be paying people a lot more. That is my problem as a founder to solve. Something that I have to work on.
Flatpak support
I don't have a lot of context, but I remember there being a lot of challenges. If you look at GitHub discussions, there is a well documented thread which Manav has maintained on what has happened and where we are right now.
Definitely we'd like to be everywhere. Distribution should not be a problem. But there are some technical things that are getting in the way. If you guys are interested, please, check out Github. We have nothing against Flatpaks. We have to choose what we focus on.
Legacy
For me, it's that when I die, I want my wife and daughter to be able to access all the photos that I've clicked, contacts in the future, Locker has all the documents that they should have access to and instructions on what to do with each of these things. Legacy ties into Locker that way. That's how we perceive the product.
Private ChatGPT alternative
No, that is not our forte.
There are companies who have raised hundreds of millions of dollars who are doing all the research and training.
Where we'd like to build expertise is on deploying these models on devices, and to get Local AI up and running when all the models are available.
I think that is the game to play. I've become better at dealing with stack envy. There is a lot of value we can create by being that company who is getting things in the hands of consumers.
I'm pretty sure there will be some LLM, I was seeing LLM a few days back, and there is llama.cpp. These things run on device. I'm sure, pretty soon in a couple of years, if not the next year we will be able to get those running on mobile.
I can totally envision Locker where we can have semantic search with your documents. "When is my tax due", "When is my fixed deposit maturing", and have a Q&A with your documents.
![Photo from AMA Photo from AMA](/static/975bcb546511ee1951b423c403d114e9/d6032/ama-2.jpg)
What's the goal of Ente?
No global domination plans.
I feel with just Photos, it's an ambitious goal in itself. To beat Google Photos or Apple photos.
In general there's this core philosphy of doing just one thing and doing one thing well, which does get in the way of spreading ourselves too thin.
At this point for the scale that we are at it's a healthy philosophy.
Who knows, in the future as we get bigger, or if someone were to throw hundreds of millions of dollars at us, it would totally make sense to let Bob go crazy and build whatever he wants to build and to setup an army of devs who can continuously support those services and maintain them.
Never say never.
Plans to translate landing page
Plans yes. Bob has been after me for a while now.
The problem with the landing page is that it's always in a state of flux. There are a lot of moving parts. It's going to take more time for things to settle down.
Docs yes, things are a little more stable there. Although I don't know if the framework we're using there supports translations out of the box. We'll have to see.
What's your exit plan?
There is no exit plan.
What are normal exit plans for companies? One is getting acquired which in our perspective is not a very sexy outcome. And the other one is going public, IPO as a company because that tries to guarantee that the company outlives the people who started it. The problem with going public is that you're then at the mercy of the market. You might be doing everything right, but the market could be unfavorable. So there are risks with going public as well, but that's one viable route.
The other one is where you stay private. There are a bunch of companies that have taken that route now. Basecamp is one of those guys.
Or you stay private and have a trust above it.
Lot of approaches.
One of my jobs in the next decade is to figure out what path makes the most sense for Ente.
The overarching idea is just to make sure there is posterity. I want my daughter to get all the photos I've clicked of her through Ente and I've to succeed at that.
So when I'm not around Ente should be around. That's the core criteria.
How many languages do you speak?
Odd question! I speak Malayalam. That's my mother tongue. The word "ente" comes from Malayalam, it means "mine".
I speak English, I speak Hindi. I can understand a bit of Tamil. I love Tamil. I feel Tamil and French are the most beautiful languages out there. They sound the best.
I know a few words in German, rather Swiss German to be more precise. I find that language very interesting. It's the most logical language I've run into in terms of how you concatinate words to make bigger things out of smaller things. Compared to English, where there is no logic.
3 languages I guess.
What was your experience working at Google? Is Google a "spyware"?
I look at the team that I worked with at Google very fondly. People are really nice, very kind folks.
In general I'm not very fond of Google, for a bunch of reasons. It's not just about privacy. I think my bigger gripe with Google is about how they pick out very smart engineers and feed them really good food and get them super comfortable to a point where in they don't leave, and they don't do anything impactful.
There's a whole meme right, "resting and vesting".
I find that very uncomfortable. There's a lot of impact humanity is missing out on because of the strategy Google has taken and what Meta has continued to take. Where you just extract out the best and put them in a room and don't let them do anything, just so they don't go out and do something else.
About the "spyware" rhetoric, I think it's a bit unfair. At their core they are an ad company. They are very open about the fact that they are an ad company. Which means they have to harvest data. The biggest moat they have against the likes of OpenAI is photos. Google will always have the best computer vision models.
Then it's about if we want to encourage "spyware" or use alternatives like Ente. This is turning into a sales pitch, but use Ente!
So I think they're doing what makes sense for them as a company and what's there in their DNA. They are no longer saying they aren't evil.
Have you always cared about online privacy?
No, I've not always cared about privacy. In the past I've worked on a bunch of consumer apps, where generally privacy is not something you prioritize.
Manav had come over on vacation at some point in Zurich, and we had a long conversation about the state of consumer data, and how it's processed. That's what got me thinking about myself and my own data. The data types that I have, and what's really close to my heart and what isn't. If I'm okay with a company harvesting some of it, all of it. "Threat modelling" as people call it now. So Google Photos was one thing that was really close to my heart.
I'm not an absolutist, different people have different threat models, and I respect that. Everybody should respect that.
I felt that as an engineer it's my responsibility to build an alternative if I can build an alternative. For my wife to use, for my parents to use.
When will national holidays be added for more countries?
I don't know, we've not been able to prioritize this. Right now there's a focus on shipping face recognition, semantic search, getting it out of the box.
From a product perspective we aren't interfering, we want the engineering team to run free and close these tasks. It will provide a lot of value to customers, by bridging these gaps.
After that we can stretch our legs and focus on iterative improvements. That's how we'd like to play this.
Plans for custom ordering in Cast, and challenges with Cast API
Custom ordering, yes. We'd love for the device to be a remote controller, so you can cast specific images instead of the whole album. We don't have an ETA yet.
In terms of challenges, James has written a really nice article in our blog. You guys should go check it out, it will do a lot more justice to the question.
Are there plans for nested albums?
I guess we've to make plans. It's something that I'm not very fond of because I feel organization can be done better with tags. But those are my personal use cases.
In general, community has been asking for this quite a bit, so we will give nested albums. There's no way around it.
What do you think of Cloaked?
No opinion, I saw their website. The website is really cool. They've a good designer, that's all I can say.
Are you religious, if so what religion?
No, I don't think I'm religious. Although I was brought up in a very conservative, religious Hindu family. That's why my name is Vishnu.
I'm not sure if you guys know of mythology. Hindu mythology is very similar to the Greek mythology. It's like the DC universe and Marvel universe. We've got our own characters and people who symbolize certain things.
My parents are very upset about the fact that I'm not as religious.
But there are certain things that have just become a habit because of the way I was brought up. Some of those habits have stuck around. I've named my daughter after the Hindu goddess Lakshmi, who stands for prosperity. Lot of those habits have stuck around not because I'm religious or because I believe in a higher power, but because there's a level of comfort and familiarity because those are things you've done as a kid.
Nuanced answer, but yeah.
When is v1 of Photos coming out?
As soon as we let Laurens get back to work.
How should I transfer 10TB of photos from Dropbox to Ente as a professional photographer?
It hurts me to say this, but if you're a professional photographer you're better off not using Ente right now because there are a lot of use cases that we don't support.
I think we'll support them next year this time. The honest answer is to give us sometime and give us some time. Check back in a year, we should be there.
How old are you?
I'm 33.
![Photo from AMA Photo from AMA](/static/aaef0ff622c4f7dd80597424142a18d5/d6032/ama-4.jpg)
Any plans to introduce more options for finding duplicate files
Yes, Bob has been experimenting with a bunch of things. He's been playing around with CLIP, he has checked out perceptual hashing. It has yielded interesting results. But we don't have anything in production yet.
Plans yes, we will implement a better way to detect similar images not just duplicates. If you've clicked 5 photos in one shot hoping that there's 1 in there that looks better than the rest.
So we will help you reduce noise for these usecases.
Does Ente offer internships or remote contracts?
Flattered that someone is asking us this.
Right now we are not hiring, unfortunately. We will setup a job board though, and we'll share it with the community so that you guys are updated as and when we're hiring.
As for now, the best way to help us would be to spread. Just tell people about Ente.
Share the landing page, just share your own photos. That's something we're working on, get sharing to a better state so more people want to use it. That way people will see Ente and that will really help us grow.
Are you considering implementing a passwordless passkey setup?
No, not right now.
I'm very new to this system of passkeys and protocols. I don't know of a way to reproducibly derive a key from which you can execute the envelope of encryption that we do. Perhaps there is a way. James or Bob might be better people to ask about this.
My bigger concern about passkeys right now is lack of portability.
If you setup this whole passwordless login system, lets say from Apple, for a specific service, and say you leave the Apple ecosystem tomorrow, there is no way to move your passkeys from Apple to elsewhere. There is no defined format for this.
I'd say be very careful when you go all in on passkeys before they figure out how stuff can be moved around.
What is your hobby?
Ente is the all time consuming hobby.
Apart from that I like listening to music quite a bit. I grew up learning Carnatic music and learning to play the violin. Although I don't really play the violin anymore. I do spend a lot of time listening to music with my daughter. Because I want her to grow up to be a musician. Live the dream that I couldn't. Hopefully she grows up to be a musician.
I do spend some time lurking on Reddit and HackerNews, although it has reduced quite a bit since I've become a dad.
I love food. I think Laurens will know this by now.
Thanks to Laurens and Bob, I've joined the gang and we go to the gym now.
We've a lot of musically inclined people at Ente now. Ashil is a great musician. Vishal also very talented. Manav also dabbles with all sorts of musical experiments. So lot of musical talent at Ente.
When will bulk tagging be supported?
We've to work on tagging as a feature in itself. Lot of folks are right now retrofitting tags into descriptions and we can do a lot better.
We're looking at tagging and nested folders as things we'd like to ship together. Don't have an ETA yet. But we'll give tags, because manual tagging is necessary to have.
When is Locker going to be available?
No clue, to be honest. Or like the meme, SOON.
We've taken a step back, because we want to do one thing, well.
We'll definitely give Locker, but not before we hit v1 of faces.
When will OCR be added?
Would like to work on that. Faces and CLIP have to be perfected first. There is some other stuff around rediscovery. Features on top of CLIP and Faces are more bang for the buck for our users. Those have to come first. Then OCR. No timelines.
Bulk actions on albums?
Same answer, after v1.
![Photo from AMA Photo from AMA](/static/7b1c9d8b22bd357c28ecae9faab87f40/d6032/ama-3.jpg)
Wrapping up
When you mentioned rediscovery, there was this question about what happens after face recognition. There are a bunch of things, video streaming.
Then sharing is a big thing. I feel that flow is broken, we go back to Signal or Whatsapp groups to share photos, and those aren't the best places to share photos because they don't encourage a conversation around photos. Where as with Ente we already have your photos, now how do we enable sharing around that? That has its own challenges. Because there's a context switch involved. You already have your conversation in Signal, now why would you switch. Or how do we bring your conversation into Ente. It's an interesting problem to work on.
Even with discovery. Most of us click photos to go back and look at it. Because those are great moments in your life. So you go back and feel joy or nostalgic about it or in general creates a sense of gratitude. There is a lot of scope for us to recreate those feelings, without you having to search for it. By stitching together information - we know who you hung out with, what you were doing with them with semantic search, and where you were and at what point. Drawing patterns from these points and creating an engine out of this. Intelligently surfacing all this. There's a lot of gratitude we can create for people, and all of this happening on the client side.
I feel there's a lot of romantic value in building this. Lot of exciting work left for us to do.
We should get going. This was supposed to be a 30m call that has gone to almost an hour.
It's Laurens' birthday tonight, in a couple of hours. It's his birthday, we got a cake there, and we've to get the party started.
I hope we'll get to do this more often. Let us know in case we can help in any way in general. We are around.
That's it, now we'll get going. Thanks a lot guys. See you around.