The main focus this past month has been on offloading some of the engineering burden so I can free myself up to focus more on project success, evangelism and business development. I’m very thankful now to have Izhar and Artur both helping me on the tech side. We’ve already made some excellent progress on a few issues. We had a perplexing server issue where the Bubble app was consuming inordinate Work Units for no apparent reason and we were able (with the help of Bubble support) to sleuth out the culprit and restore normal operation so I’m no longer paying these overages. It turns out Bubble has an undocumented “feature” where it creates a phantom user for every anonymous visitor so the logic I was using to initialize new account values was running for literally every single visitor to the site instead of just new account holders. Once I was aware of this phantom table it was an easy fix to add conditional logic to exclude those people and we saw an immediate 10x improvement in WU reduction.
Izhar has led the charge on refactoring how queries are handled across most pages converting most of those database queries into API calls (which can be cached) further saving us WU consumption as well as speeding up the page loads on all pages. He has also paginated results across all pages which contributes to faster loading times and less overwhelm on each page.
My main personal focus this past month has been spinning up a parallel fork of Problemattic that’s focused on birthing and nurturing for-profit commercial startups. This idea emerged over the course of a number of conversations where people had asked if it were possible to use the platform in this way and by the sixth one of those that I had turned down I realized this is the pull of natural demand. I grabbed the domain OpenStartup.Community and spun up a new brand ontop of the Stone Soup protocol to service this demand and run this experiment to see if we’re able to actually build startup ventures in this distributed fashion that we’ve used at our events for the impact projects. We had our first open call event for that on the 18th
and then our first work session this past Saturday:
It’s too early to say but I’m cautiously optimistic this is going to work well and address one of the biggest anchors we have which is “people need to establish the financial and temporal freedom before they have the luxury to work on pure impact projects.”
Other Wins
- Izhar implemented Sendgrid for us for email deliverability. We were hitting the max cap of 100 emails per day via the default Bubble mail provider and this allows us to exceed that cap as well as send the emails from our own domain plus track deliverability stats and improve inbox deliverability rates.
- I made it so the TypeBot-based signup flow is now the de facto signup method for all tribes that have a flow configured. If a tribe does not have a TypeBot intake flow then it falls back to the standard Bubble-based signup. This allows us to collect custom responses, build a dossier unique to each person based on the intake questions configured for the tribe they come in under and then make immediate matches. This happens in near real-time (about 1min of waiting). We can speed that up by changing the pipeline but for now I’m writing the dossiers to Gsheets as an intermediary step to ensure no downtime if/when we need to update our n8n instance.
- Intros now are generated sequentially instead of in parallel. You might think “that sounds like a step backwards processing-wise - why would you make that change?” The answer is we can now ensure that a person receives all unique introductions split into the two groups of “people they can help” and “people who can help them.” Previously when these were parallelized there was no way for us to impose the constraint that they be unique so this is an improvement by ensuring unique intros for all people.
- Izhar added an email verification step for new accounts which will be essential as we grow to mitigate the problem of spam that accompanies any blossoming community. This is done in a really elegant way which allows us to teleport the new user back to the page from which they signed up even after the detour of going through the TypeBot onboarding flow via a totally separate system.
- I added some logic to the pages that ensure Stone Soup sites running on custom domains are assigned the tribe that matches the custom domain. Prior to that it was possible that you could have cookies overwriting the tribe assignment and it was super confusing. The current implementation is a bit jhankey in how it works and you’ll occasionally see a brief glitch as a header loads with the Problemattic branding and switches to the correct tribe. Punting on this for now - it’s better than it was.
- I did this talk on Tuesday via my Meetup group here in Lisbon “Nocode Lisboa”:
This is well-worth your time to watch if you’re curious about how to use the new crop of AI builder tools to build software applications as a non-developer. It requires some patience to learn how to interface with these things but it’s well within the ability of someone non-technical who is motivated enough by the possibilities. I took a random audience suggestion for what to build and over the course of the hour I showed people how to build a little “Tinder for Dogs” application via Bolt.new and then bring it into CursorAI to flesh out features and use v0 to design screens. This is a rapidly-moving and promising space and I think there’s opportunity to create a little bootcamp type course to help onboard non-technical people to being productive with these tools without having all the computer theory background.
What I’m Thinking
- At this point I’m just grateful to have some technical help. There’s no way I could have made all the forward progress this month in spinning up the OpenStartup movement while simultaneously accomplishing all the technical wins we had without the help so I’m very hopeful that with this small team we can begin to ship quickly and make real strides in improving the platform.
- I’m also bullish on this notion of temporarily shifting emphasis towards getting a handful of commercial startups in flight and cashflow positive to prove out the OpenStartup thesis. Let’s jailbreak people from the rat race first and then ideally they’ll naturally want to contribute on the impact side once they have financial and temporal freedom.
Open Questions
- Personal time & capital allocation on the startups we have in the hopper - how to get the most advancement fastest with what we have. I like all the projects we currently have in motion but need to pick 1-2 and double-down on the most promising and then find leaders to take the others forward and support them.
- Bringing up the other 2 sides of this three-sided marketplace (investors and bounty funders). What emphasis when? I should probably begin recruitment on both of those fronts. I think it will make sense to implement some kind of syndicate functionality the way Angel List does with making it possible for any investor to be more of a capital allocator role and make it possible for others to basically piggyback his/her allocation decisions and put small amounts into pre-pre-seed stage ventures.
Next Month’s Focus
Platform
I actually believe for the first time that we’re not held back by any missing features or performance of the platform. It has all the basic ingredients (albeit still my 90’s web design UX feel 
) but I think it’s perfectly adequate in functionality for the phase where we’re at. I have Izhar refactoring some aspects that I hacked together for the sake of speed and it will be nice to pay down some of that code debt and shore up the platform we have before embarking on adding new functionality. The main missing thing I want to see us add this month is parity with the /events /bounties and /members pages so they all support the ability to enforce tribe-level filters the way that /problems and /projects do now.
Project Success
-
Message Everywhere is sooooooo close to going live. I’ve stepped out of that one a bit this last month and am still loosely steering things via the weekly standups with their interns but at this point it’s just awaiting some final testing to have confidence that the golive cutover doesn’t create issues for the comms with their current constituents. Here’s the current team of ASU EPICS interns working on the project:
-
I will be working in parallel on each of the startups shown now on OpenStartup.Community to advance those. We had ScratchMap emerge from the Hack Together hackathon which is actually a building block enabler for GeoRaces which was an idea inspired by an event we pulled off in Koh Tao back in Jan 2020. #NeverNotBoat is an idea proposed by my friends William & Megan, Behalf.bot was birthed at Startup Weekend Lisbon and I’ve been building Lunch Mixer nights as an excuse to play around and learn the AI builder tools. I would like to get LunchMixer v1 finished and deployed this month at least for our test case with the Lisbon Digital Nomads community and if it works well there then add it to the Slack Marketplace and incorporate the necessary leadgen feeders to make it be a source of talent and capital for the Stone Soup protocol.
Events
I will be traveling for the rest of December and have therefore suspended all events for us until after the new year. We will hit the ground running in early 2025 with a cadence of events aimed primarily at the commercial startup side. I’m also intending to bring up two other parallel movements once we are firing on all cylinders with OSC and collectively these four should give us broad coverage with four unique feeders (2 impact-focused and 2 commercia-focused) each with different value props and cultures bringing new people and capital into the protocol.
Content
No podcasts planned. I’m hoping I get time to turn the above video into a blog post and eventually unpack it to be a proper mini bootcamp course on Teachable or Udemy to help people come up to speed. If done right that as well could become a really interesting fountain of qualified talent for us. That course may be something I attempt over the Christmas break. I already have this outline for it:
Traffic Snapshot
The technical roadmap as always can be found here.