Three live projects and two strategic teardowns — covering fintech, food-tech, growth, and consumer product strategy.
LiveProduct
01
Pure Fitness Africa
Booking & Payment Overhaul
The app served two Lagos gym branches but had a silent Paystack failure blocking payments, no booking or class history, and users being force-logged out mid-session. I diagnosed and fixed all three, then shipped the missing core features.
Root cause
Paystack auth failed due to a misconfigured API endpoint bypassing required bank authorization steps — found via log analysis + live testing.
Key fix
Collaborated with backend to implement refresh-token auth, eliminating force-logouts. Introduced back-dated booking for in-person staff check-ins.
PM learning
"Technical depth matters. Diagnosing production failures required SQL queries, log analysis, and live testing — not just writing tickets."
0Logouts after fix
100%Payment flow restored
4Core features shipped
LiveProduct
02
FoodEase WhatsApp Bot
Operational Automation & Analytics
FoodEase was scaling fast but operations hadn't caught up. Manual Excel tracking consumed 3 hours daily, commission errors were common, and leadership had zero visibility into the payment funnel.
Nigerian university students are digitally native and forming long-term financial habits — but awareness of savings and dollar-wallet products was low due to peer trust gaps and no on-ground activation. I ran a 4-lever growth playbook across both brands simultaneously.
Activation
Large-format seminars reaching 500+ students, reframing savings from 'adult finance' to 'student smart moves'.
Acquisition
Expanded peer ambassador network for word-of-mouth at a fraction of traditional CAC.
Retention
WhatsApp community touchpoints keeping new users activated beyond sign-up — activation rate over acquisition count.
500+Students reached
25%Cowrywise campus growth
20%Cleva market share growth
PM Insights
PLG ≠ Just Marketing
Community-led growth is a product strategy. How easy it is to refer, onboard, and activate a peer is a product problem — not just a comms one.
Segment Deeply
Freshers, finalists, and working students have different jobs-to-be-done. Tailoring messaging to each cohort dramatically improved conversion.
Activation > Acquisition
Signing up 100 users matters less than getting 40 to take a second action. That needs in-person support — not another campaign.
ConceptStrategy
04
Spotify Freemium
Model Teardown — Monetisation Strategy
With 600M+ users and a ~40% free-to-paid conversion rate, Spotify runs one of the most studied freemium funnels in tech. I broke down how every friction point is deliberate — and what PMs can take from it.
Core mechanic
Deliberate friction (ads, shuffle-only, no offline) reminds users what they're missing — after the value of the catalog is already proven.
Key PM lesson
"Segment by engagement, not time. Push upgrade prompts when a user hits friction after a high-value action — that's when conversion intent peaks."
Retention insight
Spotify Wrapped + collaborative playlists create social identity — leaving means losing your musical history. Social lock-in is a retention mechanic.
~40%Free-to-paid conversion
600M+Users analysed
ConceptStrategy
05
Duolingo Streak
Gamification Deep Dive & Feature Proposal
The Duolingo streak is one of the most psychologically potent retention mechanics in consumer apps — but it's not a monetisation lever. I analysed why it works and proposed 4 features that turn it into one.
Why it works
Loss aversion + variable reward (Streak Freeze) + social proof (leaderboards) + identity formation — four psychological levers in one mechanic.
Feature proposal
Streak Milestone Certificates (LinkedIn-shareable), Partner Rewards, gated Streak Analytics for Super+, and Team Streaks for B2B.
Gamification rule
"Map every gamification mechanic to a business outcome. Engagement metrics that don't connect to revenue or retention are vanity."
500M+Downloads analysed
4Feature proposals
DevOps & Infrastructure
Building the pipeline
A 4-phase CI/CD infrastructure project transforming a manually deployed gym app into a fully automated, cloud-hosted system.
01
CI Pipeline & ContainerizationComplete
GitHub Actions workflow triggers on every push — spins up a live server environment automatically
Newman (Postman CLI) runs automated QA tests against critical endpoints including Paystack; pipeline halts on failure
Docker containers for frontend (nginx:alpine) and backend (node:18-alpine) built after tests pass
Professional README with architecture flow diagram serves as a standalone portfolio piece
02
Cloud Infrastructure via Terraform & AWSUp next
Terraform IaC files to provision EC2 instances or ECS container services — no manual console configuration
Secure VPC with configured security groups and traffic controls
Environment optimised for the Gym App's Docker containers
03
Continuous DeploymentPlanned
GitHub Actions extended to push validated Docker images to AWS ECR automatically
Auto-deploy to live AWS servers after every successful build
Zero-downtime rollout — new features ship without disrupting active users
04
Monitoring & AlertingPlanned
Server and application health tracking dashboards
Automated alerts for endpoint failures or memory spikes
Operational data for strategic decisions without digging through backend logs
Data & Machine Learning
From raw data to insight
Time-series modelling, predictive analytics, and product data systems built for real operational decisions.
🧠
PIML — LSTM Hazardous Gas Detection
Final Year Project · McPherson University · 2026
In progress
Physics-Informed ML model using LSTM networks to detect hazardous gas levels from industrial sensor time-series data. Built the full preprocessing, training, and validation pipeline in Python with TensorFlow/Keras. Currently integrating predictions into a real-time prototype monitoring system.
PythonTensorFlowKerasLSTMTime-series
98.12%Accuracy
LSTMArchitecture
📊
FoodEase Churn & Revenue Analytics
Data Engineering · Fusion Intelligence Technologies · 2026
Live
Churn-tracking system using abandoned order IDs to surface daily drop-off rates. Automated revenue grouping and commission reconciliation across multiple restaurant partners. Pipeline: Excel → Python → Google Sheets → Looker Studio → nightly PDF reports.
Open to entry level roles and internships across product management, DevOps, and data — as well as full-time opportunities after my July 2026 graduation. If your team values someone who can think in products and build in code, I'd love to connect.
Currently available for entry level roles and internships
Based in Lagos, Nigeria. Available remotely or on-site. Graduating July 2026 and actively seeking entry level roles and internships in product, technical, or data.