The problem, as it actually is
Nigeria has over 200 million people. Tens of millions of them eat out regularly. Hundreds of thousands of food vendors — bukas, mama-puts, chop houses, suya spots, fine-dining restaurants — serve them daily. But the tools that are supposed to connect diners to food still show you the same ten restaurants with inflated Google ratings left by people who don’t eat like you do.The generic algorithm does not understand jollof rice politics. It does not know that a Lagos Professional rates portion-to-price ratio above ambiance, or that a Street Food Enthusiast will forgive a plastic chair for the correct pepper. It treats every Nigerian diner the same, which means it serves nobody correctly. On the vendor side, the problem is mirrored. A buka in Surulere that makes the best egusi in Lagos will never trend on an app that prioritises paid listings and photo quality over actual food. Their customers find them by word of mouth, or not at all.
What NaijaTaste was built to do
Four things, done correctly.1. Understand how Nigerians actually eat
The Persona Engine is the cultural intelligence layer. It was not built by copying a western recommendation system. It was built by modelling three distinct Nigerian diner archetypes — the Lagos Professional, the Street Food Enthusiast, and The Aunty — each with calibrated rating tendencies, price sensitivity, vocabulary, and priorities. When you use NaijaTaste, your recommendations come from a model that knows you care about value-for-money in a way that a generic star rating cannot capture.2. Surface the restaurants that deserve to be found
Our recommendation engine draws on real restaurant data from Google Places, filtered and re-ranked through your persona. A small buka with high visit frequency and consistent positive signals ranks alongside — or above — a large chain with a marketing budget. The food decides, not the ad spend.3. Generate reviews that sound like Nigerians wrote them
The Review Simulator produces authentic Pidgin English reviews with tone labels and star ratings. “E dey sweet die but the service slow like government work.” This is not a novelty feature. It reflects how Nigerians talk about food, and it matters for any product team building review systems, training data, or customer feedback tools for the Nigerian market.4. Make the intelligence available to developers
Everything NaijaTaste does is accessible via a public REST API. No authentication required for core endpoints. A developer can integrate persona-driven restaurant recommendations or Pidgin-language review generation into their own product in under an hour.How it scales
| Concern | Path |
|---|---|
| Data freshness | Restaurant data is fetched from Google Places in real time and cached per query. Stale queries are re-fetched automatically. |
| Persona depth | Three built-in archetypes cover the majority of Nigerian dining patterns. Custom personas are composable from the base traits. |
| Language coverage | Pidgin-heavy, mixed, formal, and casual tones are supported today. Regional dialect layers are in the roadmap. |
| API scale | The REST API is stateless. It scales horizontally. No single-instance bottlenecks on read endpoints. |
Three commitments we take seriously
| Commitment | What it means |
|---|---|
| No fake data | Recommendations are built on real restaurant listings. We do not fabricate venues or ratings. |
| Persona is yours | Your taste profile is built from your own interactions. We do not sell it or share it without your knowledge. |
| The API is open | Core endpoints require no authentication. We believe food intelligence should be accessible. |
What to read next
Introduction
The Amara story. The thesis. The three sides of the platform.
How it works
Amara searches. The engine matches. The review writes itself.
Taste Personas
The three archetypes. How they shape every recommendation and review.
Developer API
Integrate NaijaTaste into your own product.