7) The Old Man and the MVP

Behind-the-scenes building Vambrace AI, a company on a mission to forge stronger relationships with users. Subscribe to follow along or visit the site here:

(typos are to make sure you’re paying attention)

Introductory Remarks

Dear Vambracers —

In last week’s post, Textual Healing, I discussed text-based analytical methodologies that I’ve already implemented and explored some of the more leading-edge techniques that I’m considering for the longer-term technical directional of the offering (based on user feedback, of course). I also put forth a crystallized vision for the company, which is to be the leading provider of commercial text analytics in the world (which also leaves some flexibility around the eventual means of business value delivery). Clarity of vision I think comes from intentional action and introspection, and I’ve done a lot of that in the past year or so (before I even formalized this endeavor) and I really feel ready to embrace the quixotic. I’m also now super into synthetic populations and neurosymbolic narrative embedding.

MVP

In today’s post, I’d like to discuss KPIs and metrics that I’ll be following as part of this initial “Private Public Beta” phase of the development of the company. I’ll also put forth my broader staging plan for community-oriented and iterative product development and then, hopefully, direct you to the MVP.

Staging Plan

I’ve previously laid out a launch plan that I recently ironed out. For the next month or so, my plan is to basically make accessible a “private public beta,” which definition I put forth as: a publicly available beta version of my product that I don’t broadly market in any way shape or form. My thinking is that I want to have a super real and production-ready version of my product that still isn’t like entirely buttoned-up, because, honestly, I’m sick of worrying about buttons and know I just need to put this thing out in the world. And then also my thinking is that what I currently have is definitely enough to go to folks in my network (if you’re reading this that means YOU!) and have them mess around with what I’ve built. I’ll also do some super light commenting and/or posting on Reddit or other more early adopterish communities to see if I can scoop up a few folks not in my network to give it a try.

I guess the goal with this phase of development is to have a very real thing that anyone can theoretically access from my website and where I can have legitimate testing sessions and get real user feedback. And then from there I can kind of dual-track: (1) better understanding my users and the use case / ICP I want to laser in on (which will eventually inform “public beta” marketing, distribution, and GTM strategy), and (2) I still need to get more comfortable with my personal development → testing → production flow, and so I basically want to work with people I have a relationship with who won’t get too mad at me if I absolutely nuke the database or do something incredibly stupid. That’s maybe fear-based thinking, but I am genuinely fearful of navigating database migrations right now, and so I do think that I want to just be moderately competent (as opposed to incompetent) before I can in good faith market this thing to strangers and cajole them to rely on my platform.

But so the good news is that, if everything goes according to plan, you should be able to access the platform like right now. More on that in a second!

The #Platty

Okay so I am pleased to report that I finally ironed out enough of the kinks and got angry enough at myself for continued delays that I have published our MVP for private public beta testing. You can find the platform HERE. Honestly, when I first got it to successfully deploy and was able to access the platform from a different device than my computer, I nearly jumped for joy. The bright moments really do make all the stress and desperation and sadness worth it (at least until the next desolation).

So, if you’re still with me, and you press on the above link, then you should see this page. I’ve bubbled the link you can press if you actually don’t really have any use for my platform but you’re still interested in seeing what we’ve built thus far and want to check it out.

We also have our normal sign-in flow (either magic link or google auth) and then our create account flow, which collects some general questions about the prospective user (3 to be exact) and then redirects them to either create an account, view a demo, or subscribe to this newsletter based on their answers.

A few more specific architectural decisions at this point:

• Authentication: As I type this, I haven’t yet ironed out the google auth for the public test domain, and so I intend to figure that out tomorrow so that I can recreate my own account and get testing with my own data, and iron out the dev → testing flow. So if the sign-in doesn’t work then just look at the demo until I the auth is back and chugging. [UPDATE: IT HAS BEEN IRONED OUT (I think)]

• Test environment: Longer-term, I intend to basically follow a local dev → testing → production flow for changes to the codebase, and I basically have one big codebase and then use GitHub for main (dev), test (test), and prod (production) branches, and then I point specific Vercel environments to those specific git branches for that all to work. For this specific MVP launch, I didn’t want to mess around with all three environments, and I still have to get in the rhythm of going from just dev → testing, as well as making all the associated database changes, and so I’m just going to ask people to use the test.vambrace.ai for now and then I’ll figure out how to migrate their data to prod when the time comes (or I’ll profusely apologize for not being able to do so).

• Demo: I’m not dug-in here, but I did decide to create a static demo experience with 4 synthetic interviews across 3 different use cases (the main demo is “discovery”) so that non-users and prospective users can pretty much see the insights we surface and the depth of our analytics without actually giving us any of their information. My hope, then, is that people will get excited at having these same insights for their businesses and that will compel them to take the ~10 minutes to upload and wait for analysis of their interviews. I also think it’s fun to let non-users see what the platform has to offer as more of a marketing type of thing (which is happening right now).

But, in general, I’m pleased with being able to actually get something out in the world, even though I know it’s still insanely rough and that it will be a slog to get feedback and that I’m not sure it can really adequately serve real pain at this point. But, hey, if the pain is significant enough then it just might work. Although, again, the point right now isn’t traction and user acquisition, it’s more ICP testing and identification, problem-solution fit, and then psychospiritual training (for me) to build some muscle just launching stuff and asking (begging) people to try it.

Goals & KPIs

Without being overly specific, because, like I said, this is really more a continuation of testing the market than anything else, there are some things that I’m actively thinking about.

1. Real data. Can I actually get people with real user interviews to use it, and, if so, then what happens? Is it at all useful? Are the insights reliable? Is it enough to get them to engage with me about what else they’d like to see or how they’d like to tweak it? Can I get even like 10 people that I know (or don’t know) to legitimately try to upload real interview data and talk to me about their experience? I think that’s the real goal for now. That would be a great starting point for streamlining the solution and probably stripping things out to create a more compelling and effective product for a specific ICP.

2. Development muscle. I mentioned this above, but I need to build out some muscle with modifying the platform and maintaining user data across multiple different environments, and I think a staged launch process is the easiest way to do that. I really stumbled my way through it this past weekend and so I feel much better than I did a week ago, but I’m still nowhere near competent enough to feel like I’d be in a position to launch like new production versions of the platty 2-3 times per week (which is the flow I want to get in once we’re at like “public beta” phase). And so I just need a little more of a ramp up there, and I think that’s okay.

3. Solution-set fluidity and mental agility. The amount of times I’ve said, “fall in love with a problem not a solution” as a VC probably is in the hundreds—and I really do mean it. The problem is that I had never actually built something (other than financial models and like “internal processes” and stuff) of my own, and so I didn’t know how easy it is to fall in love with a solution. And I really do like what I’ve built, I’m proud of the work I’ve put in, and I don’t think that’s bad. But I’m also definitely more in love with the problem I aim to address and the idea-space I intend to explore, and so I’m actively kind of remaining agile and fluid in how dramatically I change the platform. I think the core technical flow of the current solution is an adequate and good faith first try, but I’m also not married to it at all. I’ll start from scratch tomorrow if it becomes obvious that that’s what the market demands. But I just want to be hyper-cognizant of these dynamics as I actually (hopefully) get my product into people’s hands and start to synthesize their feedback (which, I’ll use my platty for).

4. Springboard for public beta. And then just to kind of really be clear about things, I’m thinking literally like 2-3 weeks in this private public beta phase and then pushing towards a real production solution that I can start soft-marketing to strangers by August (or so), and then I can further refine ICP, marketing language, and exert more intellectual energy around novel distribution channels, etc. I might end up regretting this, but I actually think I’l be able to brute force my way to distribution (even at least like 100 users or something) just because I absolutely have it in me to message and/or cold email like 200 people per day until I’ve reached a sufficiently large surface area of the universe to like get to my 100 users. I’m not saying it won’t be super difficult, but that’s already a semi well-developed muscle for me. And I think it’s fun for me to try to evangelize something and convince people to just give something a try—and so really the breakthrough for me as an individual is the ability to have something to evangelize around.

I’m also trying to remain pretty flexible in what I’m looking for right now, and am still definitely in a very experimental phase of development, so I don’t want to narrowly define goals and targeted outcomes when I don’t really know where I’m going yet. I have the North Star mission (leading provider of commercial text analytics), but I haven’t yet tested the winds, boat, current, sails, etc., and I just have to familiarize myself with those things before I can get strategic and conclusive around their use.

Looking Forward

I’m honestly so proud and relieved, really, to finally have put something out in the world. I know that it’s nowhere near ready for the big time, but everyone (and everything) has to start somewhere. If this is the worst thing I ever put out, which it probably will be, then that’s a good sign for the long-term future of the company and the platform. Now it’s all about listening to people and using that to help inform roadmap, development, marketing, distribution, and GTM.

Next week, I’ll probably reflect on anything I learn between now and then, with the benefit (hopefully) of some user interactions. And, if that doesn’t happen in sufficient volume to warrant analysis, then I’ll probably just talk more about my development process and specific hurdles that I’ve had to overcome there. I’m so thankful for Cursor and Windsurf, but the deeper I get the more certain I am that I’ll need a lead engineer within the next 6-12 months if I ever want to reach any type of meaningful scale. And there’s nothing wrong with that.

Have a great week!

Sincerely,

Luke