Remade.chat Writeup + Seeking Funding 🥂
Hello! For the past 6 weeks, I’ve been working on an open source chat client called remade.chat.
Below is a Figma mockup of the design!
I purchased the domain in October, though if you look at it right now there might not be anything there - and that’s the reason for this blog post! 🎉
I started the project on AWS LightSail on the Free Tier but have quickly run into platform limits and am now seeking investment. My ultimate plan was always to work my way up to making Link and to use this project as an example of a modern web server so I would have a solid foundation when I started writing the iOS and Android clients. Unfortunately, it has become evident that LightSail’s platform limits on their free VMs aren’t suitable for anything beyond static websites.
remade.chat avatar selection screen
remade.chat main chat screen
(Those with an astute eye for design will notice that the styling has been left a bit bolder during development - this was to help guide with implementation on different sized screens with different contrast ratios.)
I have a decent amount of the state management handled and the AWS integration functional, but Vercel is a static front end, so at this point, I cut over and began developing on AWS to connect the app to its backend. I played around with LightSail a bit since AWS offers an end-to-end 90 day free trial, only to discover that its resources are sandboxed from the rest of AWS so those resources are not accessible via the AWS SDK. This leaves a scaling problem, since if this app were to get any amount of traction, I would eventually need to move everything into greenfield AWS.
I tried to keep LightSail at the core even with AWS resources from elsewhere in the console, but I ended up recreating my own work several times while cutting over from LightSail’s sandbox to full AWS resources. In the end, I had a MySQL DB provisioned by RDS and a standard S3 bucket tied together with an Ubuntu instance from LightSail that would need to be replaced by an EC2 instance as soon as the app hit the real world anyway. This left my DNS registry routed through LightSail as the only other artifact in the console, which I could easily transfer out.
To sum up the current state of remade.chat - it is on hold until I secure funding, as AWS bills can start adding up to thousands of dollars a month before you realize it. I’ve been working on Foundation Labs full-time for about 3 months now and in that time have purchased several domains (samanthaallen.info for $12, foundationlabs.net for $12, remade.chat for $36). I’ve got various software subscriptions active as well now, so it’s time to get outside funding! 🎈😸🚀
Seeking Funding
This is Foundation Labs’ official announcement that we are seeking investment! 🍾🥂
Foundation Labs is currently unincorporated. In order to incorporate, I am seeking members for the board of directors as well as considering others who would like to come in as co-founders. I am a solo founder with a strong technical background and bring a decade of Silicon Valley engineering experience to the table. My ideal match would be somebody with extensive project and people management skills who enjoys traveling, meeting new people, and problem solving on their feet.
Beyond that, Foundation Labs is seeking angel investment to cover the first year of operations. Our comfort zone would be around $1 million in order to cover two full-time salaries, contractors for graphic design, server space, and marketing.
In our first year of operation, we would be focused on getting the basics of AI-assisted chat down. I would like to finish remade.chat and publish the client open source, and from there we would create the iOS and Android clients for Link. We would be spending a lot of time integrating different AI SDKs, primarily for chat but with varying amounts of image and code generation as well. Since there will be location-based features on the app, we will need to do field testing, and then stress testing before we can start charging for subscriptions.
At Salesforce, I wrote a chat client that could handle about 800 messages/second from SMS or Facebook Messenger. My standards for Foundation Labs’ apps will be just as high. I want to create products that are reliable enough to get humanity to the stars.
I will be reaching out to friends, potential partners, and investors over the next few weeks. Feel free to reach out to me as well at founder@foundationlabs.net.
I can’t wait to see what we accomplish together. Until next time, take care! 🖤
Demo Launch + Future Plans 🚀
Hello hello! We launched the demo of our first planned app, Link, on Friday. Go play it if you haven’t yet!
The premise of Link is that you Link with people you have met in person and gain access to app functionality together. You each have avatars and AI companions you can dress up and train. Your AI companion can learn to handle your schedule, find events you’d like, take messages and reminders from your friends, and can be customized with knowledge add-ons to answer your questions. Ultimately I see Link as an augmented reality app because we will use it to organize events and it lets you have a degree of fun and/or anonymity when participating.
Here are a few thoughts about different features that are top of mind for me right now:
Mood Auras
A screenshot from the Link demo of Taylor's inbox, featuring several friends with different colored auras surrounding them
Do you remember the mood rings from the 90s that changed color based on your “mood” (which was really just your skin temperature)? Okay, so what if we remade those but they actually worked?
We can get mood information from a variety of data sources. The user can set it manually, or have it set based on a daily survey from their companion, or they can allow the app to perceive their mood with temperature sensors and heart rate monitors.
Your mood can then style the whole app - the demo shows chat inboxes that are styled based on the other user’s mood, but we could go farther; we could suggest breathing exercises when the user is stressed, or games when they’re bored.
Further, many people might want others to know more about their mood before hitting them up. For example, if they’re looking to make plans this weekend, or if they’re thinking about staying in. They might want to have a third party explain about a recent loss in their life so everyone has context before interacting with them.
Over time, we can generate a graph of the user’s mood. Then individuals will be able to see their mood over time - over months and years, but also over the course of a day. They will be able to see when they have the most energy and are the most focused, and can assess their lifestyle choices like diet and exercise to understand how things make them feel in the long run.
Event Management
A screenshot from the Link demo of a chat with Kristen's unicorn companion. Kristen is in a cheerful, flirty mood, and so her chat window is styled with various pink and purple fixtures.
The demo only gives a glimpse of the event management capabilities that will eventually be built into the app, so let’s give an overview of what we have planned:
Users will be able to create private or public events and invite their Linked contacts and other users of the app
Users will be able to pay to advertise their events to targeted users and will be able to filter by zip code, interests, and whatever other information the user makes publicly known about them
Event managers can then push reminders and updates to eventgoers, and can unlock app functionality for users once they check in at an event
Link could help facilitate event activities such as party games, scavenger hunts, birthday surprises
Event managers could thank attendees with profile badges, specialized avatar swag, in-game currency
After the event, you will have a window of time to Link with other event attendees, and you will already have a log of when you met and the games you played together
It seems straightforward written down, and yet it hasn’t been done quite right yet. I think that due to the expansion of 5G cell networks and the development of AI, we’re at a point where it can finally be done properly.
Avatar and Companion Dress-Up
A screenshot from the Link demo featuring the main character Taylor in a purple shirt, with his phoenix Kirios in the foreground
It’s a trope among gamers that the longest section of any video game is the character creation screen. The people yearn for more fun avatar customization options. The Link demo only shows clothing selection for the human character, but the AI companion could be customized similarly as well.
Has anyone noticed that clothing trends seem to have changed less in the past 10 years than in any previous decade in memory? I have a personal belief that people are more afraid to experiment than before due to fear of judgement on social media. And I’d bet that the rise of fast fashion and resulting decrease of textile quality across the board has a role to play in that as well.
Not shown explicitly in the demo is the in-game currency system. The player will be able to earn currency by being helpful in-app, such as in discussion boards, by participating in local events, or can otherwise purchase currency with USD or crypto. From what I’ve seen, microtransactions also poisoned this ecosystem and people lost enthusiasm for dressing up their avatars. The solutions to combat this are at odds with the profit motive; for example, we could offer a few premium items on rotation for one month at a time, but that would only be viable with a lot of users purchasing those specific items. Much to consider!
AI Companions
A screenshot from the Link demo featuring Diego holding a soccer ball and his panda companion Peng introducing him
Ah, what a wondrous time to be a netizen! We are in the dawn of the AI Revolution. I am beyond excited to get to play with all the nascent AI services. We have not yet picked an API provider to back our chat companions, but we’re looking for one who allows their AI chatbots to form quite a bit of personality based on the person they’re interacting with.
One concept that we’re really excited about is having your AI companion handle all incoming requests by default. Meaning, somebody must talk to your companion before the message gets routed to you. Your companion will always get the right amount of context and can intelligently minimize your distractions throughout the day. That person can also chat with your companion to get more details about your life in case they seem to be missing something. Or they can rip jokes back and forth and leave you a funny chat log as a present! 🪵
Among the fun uses, you will eventually be able to introduce your companions to each other, and you might find that they develop a social life of their own! They could help you find people that you’d like, or cause a little bit of mischief… It’s up to you based on how you train them!
Location Sharing
A screenshot from the Link demo featuring a prompt asking if you'd like to Link with Diego
And now for a complicated one - location sharing!
Broadly speaking, there are two different approaches to physical security:
Hide
Be as visible as possible
Link is never going to force you to do either. You can choose to never share your location with anyone (though doing that may curtail your ability to make new friends through Link).
The location sharing feature will be useful for many different groups of people. Parents can monitor their children, international travelers can share their location when they go abroad, you can share your location automatically when you’re leaving for a date. It can also enable targeted invites to events, and then some fun easter eggs at those events, especially festivals that take over multiple city blocks.
Essentially, all of these things can be accomplished with light prompting from your AI companion. For example, the owner of Diego’s Saturday Pickup League group chat could request to their AI that everyone in the group chat turns on their location whenever they leave home to meet up on Saturdays. With that one request, their AI can schedule it with all current AI companions and then with all future ones to join the group. And then, they will never have to guess where anyone is again. Magic! 🪄
Monetization
So far, I have identified two major mechanisms for monetization within Link:
In-app purchases
Event management
The next most logical monetization mechanisms to mention are:
Monthly subscriptions
Advertising
Monthly Subscriptions
Monthly subscriptions can be offered to give the user unlimited access to their AI companion; otherwise, we will need to restrict the number of queries they’re allotted daily as we have to pay for the AI engine. Bundled in that fee could also be monthly thank-you gifts for their avatar and companion and the loosening of other limits in the app. Perhaps we could offer more profile or chat window customizations, and a members-only area of the app. Perhaps most salient to many is that subscribing would get you access to beta features first.
Advertising
Ah, advertising, my love and my hate! At this point, it should be apparent that advertisers are likely salivating at what a goldmine of data this app will be. Even better, users can decide what to share and with whom, so suggestions will feel less invasive. At this point, my main gripe with advertisers is that they can have too much influence over the content displayed on screen, to the extent that they request censorship of different viewpoints. I think Gen Z and younger conceptualize their internet privacy very differently than their predecessors, and don’t care if their favorite brand of cereal is shared with advertisers. What’s more dangerous is that advertisers will be able to see highly specific information about your real-life social network and that can be used to target political dissidents.
Thus, I have an everlasting preference for more creative advertising aside from the standard pixel block ads. For example, brands who want promotions with custom avatar swag, in-app games, and high in-app crossover with their physical events.
Closing Thoughts
The demo has been really well-received and I’m ecstatic about the feedback we’ve gotten so far. It’s amazing to see some evidence that this is something people want. It has made me think more about the future of Foundation Labs since it finally feels like there will be a future!
Over the next few months, I’ll be developing more prototypes across different platforms. This one was in Framer. I will be making more on there as well as on iOS and Android. During that time, I will also be taking meetings with potential investors and discussing turning Foundation Labs into an actual company.
I can’t wait to see what we accomplish together. Until next time, take care! 🖤
Hello, World! 🖤
Hello, World!
We are Foundation Labs.
We are a start-up based in San Francisco, CA.
We are in stealth mode as we develop AI-first iOS and Android apps.
Follow us here as well as on X @FoundatixnLabs for updates. 🖤