← Back

Co-founder & CPO · Akaani

Building the first AI-powered nutrition platform for Nigerian meals

2,000+ signups · ~200 DAUs · 42% DAU lift post-Lu · iOS & Android · React Native

useakaani.com ↗

My Role

I co-founded Akaani and own the product function end-to-end: vision, roadmap, growth strategy, and a team of a product designer and product marketer. My co-founder owns engineering. Every product decision is mine.

Where this came from

This product is personal. I grew up cooking and eating Nigerian food, and when I got serious about fitness, I hit a wall: none of the nutrition tools worked for the food I actually eat. MyFitnessPal didn't have the meals. Generic LLMs hallucinated macros. My Nigerian roots, my love of cooking, and my fitness journey converged into one clear problem: people who eat African cuisine are invisible in nutrition tech. Akaani is the answer.

Problem

Macro tracking tools don't work for Nigerian food because the data doesn't exist in mainstream nutrition databases. Millions of people eating jollof rice, egusi, suya, and pounded yam have no reliable way to understand what they're eating. Everyday food decisions are hard enough: time, culture, preference, and nutrition all pull in different directions at once. Without the right data, no tool can help.

What makes it different

A proprietary dataset of 200+ authentic Nigerian meals: macro breakdowns, cultural variants, ingredient-level tagging, built before LLMs existed. No general-purpose model has this context. It's our moat and what makes Lu accurate where every other tool fails.

System Design

Layered deliberately: structured data first, intelligence on top.

  • Data layer: 200+ Nigerian meals with macro data, ingredient tagging, and cultural variants, collected before LLMs existed, now a durable moat
  • Rule-based filtering: dietary restrictions, allergies, and nutrition goals applied deterministically
  • LLM layer (GPT-3.5): conversational flexibility grounded in the structured data, not raw generation
  • Habit loop: weekly meal plan + grocery list, giving users a reason to return every week

AI Assistant: Lu

Lu is the primary interface. Onboarding is conversational. Lu asks about meal preferences, dietary restrictions, nutrition goals, and allergies, then uses that profile to generate a personalized weekly meal plan and grocery list. Users don't navigate menus; they talk to Lu. After launching Lu, DAU grew by 42%.

Growth Strategy

I own the full growth funnel: acquisition, activation, retention, and expansion.

Acquisition

Referral loops: Built in-app referral prompts that encourage users to share with friends who eat Nigerian food. Word of mouth is our strongest channel because the problem is deeply personal and cultural.

SEO: Optimized for long-tail searches like "Nigerian food macro tracker" and "jollof rice calories." We rank for queries no other app targets because we own the niche.

Social media: Organic content focused on Nigerian food culture, nutrition myths, and meal prep. Authenticity drives engagement in this community.

Result: 2,000+ signups through organic channels. No paid acquisition yet.

Activation

Lu onboarding is the activation moment. Instead of dropping users into a complex UI, Lu asks conversational questions: What do you like to eat? Any dietary restrictions? Nutrition goals? This reduces friction and gets users to their first meal plan in under 2 minutes.

Result: 42% DAU increase after introducing Lu. Conversational onboarding drives activation.

Retention

Weekly meal plan habit loop.Users generate a meal plan and grocery list every week. To reduce friction, they can generate next week's plan mid-week, so they're never caught unprepared. The weekly cadence creates a recurring reason to open the app.

Result: ~200 DAUs. Retention lives or dies on habit formation, and the weekly loop is working.

Expansion

Currently free. Building toward a paid subscription tier for premium features: advanced meal customization, nutrition coaching, and expanded meal database. The free tier validates the core value; paid will unlock deeper personalization.

Product Decisions

Finding our moat: from everything food app to AI nutrition platform

We struggled to identify our competitive advantage for a long time. We started as a marketplace, then pivoted to a recipe platform, then tried to be an "everything food app." We kept pivoting because nothing felt defensible. At some point, we were just collecting data on Nigerian meals without a clear product direction.

The breakthrough came when I paid closer attention to African nutrition. I realized: no mainstream nutrition tool works for Nigerian food. MyFitnessPal doesn't have the meals. Generic LLMs hallucinate macros. But we had collected 200+ authentic Nigerian meals with full macro breakdowns, working with Nigerian dietitians for accurate nutritional data, before LLMs were widely available. When the AI wave hit, we realized what we had: a proprietary dataset no general-purpose model had been trained on.

That pivot from "everything food app" to "AI nutrition platform for Nigerian meals" was the founding insight of Akaani.

What I learned: Our data advantage wasn't engineered. It was earned through years of unfocused data collection. Sometimes the moat reveals itself after you've already built it. The key was recognizing it when it appeared.

Introduced Lu instead of expanding static UI

Rather than adding more filters and screens, I replaced the navigation model with a conversational assistant, reducing friction and driving a 42% DAU increase.

Tradeoff: Required careful prompt design to keep Lu grounded and trustworthy.

Designed a habit loop around weekly meal planning

Retention lives or dies on habit formation. The weekly meal plan and grocery list give users a concrete reason to open the app every week. To reduce planning friction, users can generate next week's meals and grocery list mid-week, so they're never caught unprepared at the start of a new week.

Tradeoff: Weekly cadence requires users to engage proactively. The mid-week generation window directly addresses this by turning a potential churn point into a second weekly touchpoint.

Impact

  • First app to provide macro tracking for Nigerian meals with 200+ meals in the database
  • 2,000+ signups through organic acquisition (referral, SEO, social media)
  • ~200 DAUs driven by weekly meal plan habit loop
  • 42% DAU increase after introducing Lu (conversational onboarding as activation moment)
  • Building toward paid subscription tier for premium features

If I Were Building Today

I'd move Lu toward proactive intelligence, surfacing meal suggestions based on time of day, past choices, and seasonal availability before the user asks. The data foundation is there. The next step is making the system anticipate rather than respond.