core fragments: visions of a greater impact at astra

This is just an early stage ramble to help me organize my thoughts: no technical research, idea of how to execute, or timeline for making it happen.

Part of what is an inevitable reality at Astra is resourcefulness. The vision is that it’s not just a set of software focused on one issue, but interconnected software that serves people through the issues they’ll face among their lives.

Oftentimes, people going through hard experiences will have intersectional issues: domestic violence victims may experience higher financial instability, and more mental health problems. People who have long-term disabilities might experience mental health issues, and issues having the energy to search for resources.

It’s my belief that to truly build software that impacts people’s lives, it’s not enough to relieve one area of it: it would be better to have a set number of software that addresses common issues, and help people combine them to fit their situation.

But, that’s hard to execute in the real world. Donors don’t want to give to multiple causes, they want to give to one (which Astra found out back in 2018). Foundations don’t really like to support risky models of providing help, especially to young founders like myself, and prefer to support proven founders and research based programs. We’ve had a lot of donors who say they’re ready to step up to the plate, and pull donations that would have been life changing for the company (like a 500k capacity building donation that never materialized after committing and then ghosting).

At some point, you say “fuck the funding, let’s try to build what we can nevertheless“.

Astra’s been moving into a position where we save resources by strategically investing in templatization, like Aureus which standardizes our front-end design system with safety and accessibility functionality.

I’m thinking of how we can continue not only to reuse design, code, but also product development features to improve the quality of all resources, reduce the amount of time to develop, and to drive traffic to the other resources as a bit of a ‘Costco Free Sample’ move.

Everine is a giant database of digitally accessible help across a variety of different issues. So, the ‘core fragment’ of it is resources for a specific issue. It would be incredibly easy to integrate it as a side page that says “Pulled from Everine”, and to just pull from the Everine database.

Now we can imagine integrating that into Aetheria: an app for people going through mental illnesses, and Verena: an app for people in dangerous situations. Aetheria could have integrations into free therapy funds, and Verena could have resources for those going through abuse and other situations.

Something that we’re having to think about is the product balance of the resources. We want to keep lean products for smaller development cycles, less bugs, and higher quality core products.

However, because we’re broke as fuck, we want to create partner software to our resources that enable us to generate revenue. Like building a therapist’s platform for Aetheria that would allow therapists to assign specific tools.

It creates the paradox of: ’how do we balance the need for a lean product with needing it to be fully featured / developed enough to justify the value of an enterprise tool?’

I think we have a potential way to address it through core fragments. Each of the resources has a clear utility that can be applied to at least one other resource.

For example, a scrapped feature that was going to come out with this version of Verena is a tool called ‘Guided Cooldown’, which was just Aetheria’s tools meant for people who have been through a traumatic situation.

Eventually, I want to address the current and soon to be worsened caregiver crisis and disability justice movement by creating an app that helps people with long term disabilities gain independence, and for caregivers to have an assistant that reduces their burnout.

There’s a lot of potential integrations with Verena’s SMS alert system to have people alert their family members in the event of a fall, if they need food, or so on and so forth.

I’m not sure of the execution, or how this would be technically implemented. Maybe the UI components would be standardized in the front-end, and it would be a database full of fragments that get pulled into the interface.

But, I think this is a good step towards the future we want to build more. Hopefully I’ll update as I figure more out, and or Jeff Bezos gets sick of Blue Origin and leaves his human-life improving budget to Astra Labs. A girl in a jail trailer can dream.