I started We The Gente, a nonprofit helping people achieve their life goals, whether it’s education, a trade, or a dream still taking shape. We help uncover talents people may overlook. Our work is about getting people where they want to go and helping them see their own worth.
There’s a deeper reason, too. My parents came from Mexico and became citizens, but they and many others still associate “American” with being white. I was born here, served in the military, received an education, and am raising my family. We The Gente is built on that sense of belonging. We build ourselves and take up our space.
That’s the foundation. We The Gente
When the surge hit
In late 2025, we had just held our first board meeting. Our entire track record as an organization was one thing. We got a noise-canceling headset for a school kid who needed quiet to study. That was it. That was the whole operation.
Then the ICE surge moved into Minnesota and rolled into 2026. People in our community stopped leaving their homes. They weren’t grocery shopping. They weren’t going to work, so bills ran late, and eviction risk climbed. We had no infrastructure, no playbook, and no money. We saw the need and started moving groceries anyway.
We were careful about who we let help. In hindsight, opening the doors wider would have helped. I prioritized privacy because everyone felt vulnerable, and we only worked with those we knew personally.
The delivery system
Day one, the whole thing ran on sticky notes and through messaging across multiple mediums, it got crazy super fast! The first real problem was how to coordinate deliveries without moving families’ personal information more than necessary. So we built around a simple privacy principle. Anyone handling a delivery worked from the minimum required to do the job; full details were provided only at the moment of delivery, and nothing persisted afterward. No information sitting around longer than it had to.
That meant sticky notes with limited information and an identifier for the family and home. Privacy chats and auto deleted messages for outgoing. The incoming, however, was super messy, we are talking texts on one phone being handed off all the time between three people, whatsapp messages, facebook messages, tiktok messages, everything and everywhere. We had to make sure we were where people needed us to be.
That quickly evolved into a real tool that made a significant difference. I used AI to build it out. Families fill out a form on the website, the data plots them on a map so drivers can batch deliveries by area, and the system prints shipping labels. The label shows the packer how many people are in the household, how many kids there are, and what they actually need. If a packer saw kids on a label, they might throw in extra snacks. That part stayed human, left to their judgment. The tool handled the logistics so the people could handle the care.
We delivered groceries to over 700 unique families in that window. Some families got more than one delivery.
From groceries to grants
The money moved fast. Donations flooded in, then stopped, and for a stretch we ran on our own money. Getting connected with Stand With Minnesota brought a real influx that kept food moving. We got through it with money still left over, and the question changed. We were a year old, a board of three, two of us local and one out of state. We had a nonprofit in good standing and a community that needed more than emergency groceries. So how do we keep growing? How do we keep helping after the crisis cools?
We’ve since paused the pantry as it ran during the surge. I still want to do pantry work, but it can’t stay “submit a request, and we drop it off.” That was right when people couldn’t leave the house. The next version means sitting with families, understanding where they’re actually hurting, and meeting that specific gap. Sometimes the gap is food. Often it’s rent or a bill. Then help them move towards their family’s defined goal.
The shift between going full speed, getting groceries to families, and hearing their stories to near quiet is hard to move between. Though it is nice for people not to have to worry, the shift is indescribable.
Which led me to the real problem. How do we grow? Where do we go from here? Finding it is the work.
Building the grant tool
My day job is site reliability engineering, so this is where my two worlds met. I’d been hearing a lot about OpenClaw and what agents could do, scouring the web on your behalf. I’d already used AI to build the delivery tooling, so I carried that forward.
I stood the system up on a single Virtual Private Server (VPS) to manage everything simply and reliably. I did the security by hand first, locking down ports, SSH-only access, and the basic hardening, then used Claude Code to stand up OpenClaw on the box. I wired it to Tavily and news APIs so the agents could pull publicly available grants, funding news, and partnership opportunities. Everything landed in SQLite, with a dashboard on top, and an AI layer grades each item on a scale of 1 to 5 for how well it fits what we do.
I built it as a monolith on purpose. The tool is observational. I log in and read what it’s gathered. So I kept every piece of personal data off that box. If someone broke into that server, all they’d walk away with is publicly available information, just better organized. When you’re a one-person operation, the cheapest security control is not having sensitive data on the server at all.
Then the tool started paying off, and I learned how much historical data matters. A grant that exists this year almost certainly exists next year. If I miss one this cycle, I want the system to hand it back to me next cycle. A single point of failure was fine for an experiment, but not for something the org now relies on. So I migrated. The database moved from SQLite to Supabase Postgres, the front end moved to Vercel with the code archived in GitHub, and the agent workloads stayed on the VPS, which still holds nothing sensitive. The next thing I’m evaluating is Hermes, which looks like the stronger agent framework from the pros and cons I’ve run so far.
Who I’m building for
The tool does more than chase grants now. I’ve added data points for helping people with criminal records get a foothold and for veterans. I’m a veteran, and from what I’ve seen, Hispanic veterans don’t use the resources available to them, which feeds into homelessness and/or worse. A lot of what gets treated as normal in our culture, the expectations of how a man is supposed to act, the drinking, can quietly mask PTSD. I want to make sure we can reach those men.
I came up rough, like many others. So when searching for grants and programs, I focus on helping younger versions of myself, knowing there are many with similar needs.
I also started building a funder side. It pulls 990 filings and ranks funders by what they’ve given in the past, matched against what we do and who we serve. It tells me which funders are most likely to fund We The Gente, so I’m not guessing. Between that and the grant news, I now have data on three things at once. The grants and programs out there, the funders most likely to back us, and new nonprofits entering the space.
Tech first, on purpose
We The Gente is tech-first by design. In our communities, we default to manual work. We don’t push toward tech, or STEM, or office careers, for the most part. I want to change that, position our people so we’re not working ourselves into the ground, and help ourselves and the generations after us. So that in 90 years there’s something to show for it.
The rest of the stack serves that. We run omni-channel chat, so we’re reachable wherever people already are: WhatsApp, Facebook Messenger, plus email. Automation to handle a lot of the backend. VoIP covers calls. None of it is exotic. All of it is meant to let a very small team move real weight.
We still have a lot to learn and a lot left to do. Having the data and funds is one thing; after that is piecing it all together and actually running programs that help people. Hopefully this shed some light into my thinking and setup. If you have any questions, feel free to drop me an email.