Melanin Map

Melanin Map is the consumer face of the Djembe AI thesis: that culturally grounded software, built voice-first and rooted in the communities it serves, is a category that doesn’t exist yet at scale. You open the app, you find stories, landmarks, businesses, and events curated by the community. Big Mama lives inside as the voice-first guide.

This is the first product I shipped under the Bingo Codes founder track. It’s live on the App Store today. Everything I write about voice agents, RAG over cultural data, and consumer voice UX traces back through what we built here.

What’s actually in the app

Four surfaces, one map. Stories give you Black history and culture in context, not as a quiz. Landmarks and Black-owned businesses put the places that matter on a map you can actually use. Events surface what is happening near you this week. And Big Mama is the voice-first guide threaded through all of it, so you can ask the way you would ask a person instead of hunting through menus. It is shipped and in people’s hands, which means the feedback loop is real, not hypothetical.

Discovery is the whole point

Discovery is not a nice-to-have, it is economic infrastructure. If people cannot find a business, the business loses the opportunity, full stop. Generic search runs on paid placement and popularity metrics, and that quietly pushes the small, community-rooted businesses to the bottom, the exact ones a neighborhood depends on. Melanin Map exists to flip that. Instead of typing a query into a box that does not understand your context, you ask for a Black-owned restaurant near you, a barber with weekend slots, a family-friendly event this weekend, and you get an answer that is current, relevant, and trustworthy. Making discovery conversational is not a gimmick, it lowers the barrier for the people generic search serves worst, including the elders who hold the most community knowledge and want to ask out loud.

Built with, not just for

The hardest constraint in this product is not technical, it is ethical, and it is load-bearing. AI for a community becomes extractive the moment it takes data, attention, or cultural value without returning ownership, benefit, or control. So culturally grounded cannot just mean the interface sounds familiar. The value has to flow back to the people and the businesses the system represents. That shows up in the product decisions that matter: business owners owning and correcting their own profiles, users able to flag a bad recommendation, cultural knowledge that carries its sources instead of getting scraped and flattened, a system that says so when it is unsure instead of overclaiming. Some of that is live and some of it is the bar I am building toward, and I would rather name the bar than pretend I have already cleared all of it. Building it in public is part of the accountability, it invites the scrutiny and correction that keep the thing honest. I wrote the longer argument up in Building AI for Black Communities and SMBs.

Where it is

Melanin Map is live on the App Store now, and it is where Big Mama lives as the voice-first guide, so the two grow together. The voice engineering, the RAG over cultural data, the consumer voice UX, all of it gets written up as I go in the Voice AI Engineering series. If you are building in this space, or you care about who the next decade of AI is actually built for, find me on X.