Pivoting from 3+ Yrs Data Analysis to Applied Data Science & ML

Data Analyst Pivoting to Data Science & ML

Bridging 3+ years of data analysis, SQL/Python analytics, and Web3 research with machine learning, anomaly detection, and predictive modeling.

Results-driven Data Scientist and Researcher (with 3+ years in Data Analysis at Flipside Crypto & MetricsDAO). Specializing in exploratory data analysis (EDA), predictive modeling, anomaly detection (MEV/sensor analytics), and database-grounded AI tools. Recently advanced data science and statistical modeling through the ALX Data Science Program.

Scikit-LearnPredictive ModelingAnomaly DetectionAdvanced SQLPython (Pandas/NumPy)Power BIPydanticAI / GeminiSignal Processing (FFT)Flipside Crypto
3+ Years
Data Analytics & Web3 Research
ALX DS Cohort
Advanced Machine Learning & Modeling
15 Essays
Data & ML Technical Publications
Data Science Pipelines & Machine Learning Architectures

Featured Projects & Pipelines

Machine learning models, exploratory data pipelines, Web3 analytics engines, and database-grounded AI tools.

Data ScienceOngoing

Personal Fitness Tracker

Sensor-Based Physical Activity Classification & Motion Pipeline

A data science processing pipeline and predictive modeling framework to classify physical exercises and track movement patterns using raw accelerometer and gyroscope sensor data from wearable devices.

  • Multi-frequency sensor data alignment (Accelerometer at 12.5Hz, Gyroscope at 25Hz)
  • Time & Frequency domain feature extraction using Fast Fourier Transform (FFT)
PythonuvJupyterpandasnumpy+3
AI & MLDeployed

SME Biz Analyst

Mobile-First Accounting & Grounded AI Financial Intelligence

An audit-proof mobile-first accounting and business intelligence application for small businesses in Nigeria. Combines dynamic transaction ledgering, automated night-end Telegram financial PDF reports, and a non-hallucinating database-grounded AI analyst.

  • Grounded AI assistant using PydanticAI to execute structured database queries over business tables (zero-hallucination finance)
  • Hybrid CRUD Architecture: Frontend directly queries Supabase with Row Level Security, keeping backend footprint lightweight
ReactTypeScriptViteTailwind CSSshadcn/ui+9
Web3 & DeFiDeployed

Pharos Developer Onboarder

Agentic Skill & Developer Toolkit for Pharos Network

An interactive agent-driven onboarding assistant and developer toolkit fast-tracking integration with Pharos Network (a parallel-execution blockchain featuring ultra-high throughput and sub-second finality).

  • System Diagnostics & Environment Scanner for Node.js, npm, yarn, git, and Foundry
  • Multi-toolchain scaffolding templates for both Foundry and Hardhat workflows
SolidityJavaScript/ES6+PythonFoundryHardhat+5
Web3 & DeFiDeployed

Nexus-Sentry V2

AI-Augmented DeFi Execution Intelligence Platform

A cloud-native execution firewall built natively for the X Layer DeFi ecosystem. Intercepts trade intents, executes live multi-path routing simulations, and leverages LLM reasoning to rewrite transactions into capital-efficient paths to prevent slippage and capital loss.

  • Real-time execution firewall bridging user intent and on-chain execution
  • Multi-path routing evaluation: Direct, TWAP Split, and CEX-DEX loops
Next.jsVercelFastAPIGCP Cloud RunGemini Flash+3
Web3 & DeFiDeployed & Paused

Zcash Explorer

Privacy-First Hybrid Blockchain Intelligence Platform

A next-generation blockchain explorer built for the Zcash Data & Analytics hackathon track. Bridges public analytical transparency with zero-knowledge shielded transaction privacy via a decoupled web and local desktop architecture.

  • Decoupled Hybrid Architecture: Public data rendered via Next.js web app, shielded Zcash decryption isolated to local Rust desktop app
  • Interactive money flow graph network rendered via React Flow
Next.js 15Tailwind CSS v4Bitquery GraphQL APIRechartsReact Flow+1
Published Research & Technical Writing (15)

Essays & Publications

Deep dives into crypto-economics, stablecoin yield moats, MLOps, regression math, and software architecture.

Crypto-Economics7 min read

Revenue Defensibility for Stablecoins

Crypto-Economics, DeFi Market Microstructure, Product Strategy

Analyzes competitive landscapes and long-term business viability of stablecoin protocols. Breaks down how issuers survive commoditization and build defensible yield revenue models by applying classical economic moat frameworks to decentralized networks.

Takeaway: A stablecoin's true barrier to entry is rarely its technical design; it's the sheer depth of its integrated liquidity network.
Medium
MLOps & DevTools5 min read

Using Git in Antigravity IDE

MLOps, Version Control, Model Deployment, Developer Workflows

Explores developer experience and version control patterns tailored for the modern machine learning lifecycle. Focuses on isolating, debugging, and maintaining rigorous version history when transitioning models into cloud IDE runtime environments.

Takeaway: Shifts version control from a static repository tool to an active diagnostic ledger.
Medium
MLOps & DevTools4 min read

Introduction to Antigravity IDE

Software Engineering, Development Environments, Cloud Orchestration

Introduces core architecture and setup configurations of cloud-native development environments. Focuses on easing environmental initialization friction and providing data science teams with optimized compute clusters.

Takeaway: Eliminates environmental friction bogs when scaling scripts into multi-module systems.
Medium
Machine Learning6 min read

A Beginner's Guide to Pandas Library

Data Science, Data Manipulation, Data Engineering

A practical breakdown of pandas for tabular data manipulation in Python. Emphasizes developing structural intuition for memory-efficient data cleaning, series evaluation, and indexing over raw flat files.

Takeaway: Moves past simple syntax to build intuition for memory structures under the hood.
Medium
Python Architecture4 min read

A Beginner's Guide to FOR Loop in Python

Python Programming, Control Flow, Algorithms

Deconstructs iterative logic patterns using Python for loops. Covers sequence iteration, dictionary lookups, and algorithmic repetition that drives data processing scripts.

Takeaway: Iterative clarity prevents early algorithmic complexity traps in data scripts.
Medium
Python Architecture6 min read

A Crash Course in Object Oriented Programming (OOP) in Python

Software Design, OOP, Python Architecture

Focuses on transitioning from linear scripting to modular, scalable Object-Oriented Programming (OOP) patterns in Python. Examines class blueprints, state encapsulation, and inheritance trees.

Takeaway: Encapsulating state prevents global variable leakage in complex pipelines.
Medium
Machine Learning5 min read

Introduction to Scikit-Learn Library

Machine Learning, Scikit-Learn, Predictive Analytics

Analyzes scikit-learn as the foundational toolkit for classical machine learning. Details standard API designs, data preprocessing estimators, and pipeline hygiene.

Takeaway: Pipeline encapsulation guarantees feature transformation hygiene during cross-validation.
Medium
Applied Statistics6 min read

Decoding the Data: Mastering Multiple Linear Regression

Applied Statistics, Predictive Modeling, Regression

Details the statistical math and practical application of Multiple Linear Regression. Explores weight relationships between multiple independent features and a target variable.

Takeaway: Feature interactions require checking variance inflation factors before deploying weights.
Medium
Machine Learning5 min read

Machine Learning Income and Happiness Analysis

Exploratory Data Analysis, Data Storytelling, Applied ML

A case study using exploratory data analysis and regression primitives to evaluate correlation between socioeconomic income metrics and behavioral happiness outcomes.

Takeaway: Data intuition derived from exploratory plotting yields higher model clarity than immediate complex neural nets.
Medium
MLOps & DevTools5 min read

Model Accuracy & Performance

Model Validation, MLOps, Statistical Evaluation

Breaks down evaluation metrics for predictive frameworks, exploring how to select validation metrics (MSE, RMSE, R-squared) that align with real business targets rather than raw accuracy.

Takeaway: Poor metric selection leads to unexpected production failures.
Medium
Applied Statistics6 min read

Multiple Linear Regression: Matrix Math & Vectorization

Linear Algebra, Optimization, Core Algorithms

Focuses on building out matrix math, gradient descent calculations, and vector tracking equations behind multi-feature regression models.

Takeaway: Matrix dot products provide massive performance leaps over manual looping in Python.
Medium
Applied Statistics4 min read

Linear Regression: Supervised Learning Primitives

Supervised Learning, Parametric Modeling, Data Foundations

Acts as an entry point to supervised learning, tracking how simple linear models translate single feature patterns into continuous numeric predictions.

Takeaway: Simple white-box baseline models should always precede complex neural architectures.
Medium
Applied Statistics5 min read

Least Squares Regression: Calculus & Loss Minimization

Mathematical Optimization, Loss Functions, Calculus

A mathematical deep dive isolating Ordinary Least Squares (OLS) as an optimization criterion. Details the exact derivative process used to minimize squared residuals.

Takeaway: Analytical OLS solvers give exact weight solutions without requiring step-wise gradient descent iterations.
Medium
Machine Learning5 min read

Introduction to Predictive Modeling Lifecycles

Predictive Engineering, ML Lifecycles, Data Strategy

Tracks the full lifecycle of predictive engineering, mapping projects from initial raw data ingestion through feature curation to deployment.

Takeaway: End-to-end ML scoping prevents mismatch between model outputs and business requirements.
Medium
Machine Learning5 min read

Machine Learning Paradigms: Supervised, Unsupervised & RL

Artificial Intelligence, Core Paradigms, Systems Overview

The foundational post of the portfolio series, mapping out the machine learning landscape. Differentiates when to deploy labeled predictive systems vs unlabeled clustering structures.

Takeaway: Choosing the correct learning paradigm upfront dictates the required data architecture.
Medium
Career Pivot Journey

Professional Experience & Education

3+ years in data analytics and Web3 research, continuously evolving into advanced data science and machine learning.

Work Experience

Data Analyst / Researcher

Nov 2023 – June 2025

Flipside Crypto (Remote)

Performed deep-dive exploratory data analysis (EDA) using SQL and Python to identify protocol performance trends and inform strategic decision-making.

  • Engineered comparative analysis frameworks for Web3 governance systems and EIP-4844 L2 gas fee dynamics
  • Investigated MEV bot anomaly detection on Base L2 using statistical fee distribution models
  • Managed automated campaign performance dashboards monitoring real-time operational KPIs

Web3 Data Analyst

Aug 2022 – Nov 2023

MetricsDAO (Remote)

Transformed raw multi-chain data (Solana, Flow, Terra2) into actionable analytical insights and post-collapse ecosystem metrics.

  • Built comprehensive KPI dashboards for ecosystem recovery tracking
  • Facilitated technical protocol data cleaning and pipeline integrations for MetricsDAO platforms

Blockchain Data Analyst Intern

June 2022 – Aug 2022

Flipside Crypto (Remote)

Collected and transformed DeFi data streams into real-time analytical dashboards.

  • Built custom ecosystem analytics dashboard for Sweat (NEAR)

Education & Specialized Training

ALX Data Science Program

Jan 2025 – April 2026

ALX Africa

Focus: Machine Learning, Statistical Analysis, Scikit-learn, Feature Engineering, Signal Processing & Predictive Modeling.

ALX Data Analytics Program

July 2024 – Dec 2024

ALX Africa

Focus: SQL, Python (Pandas/NumPy), Visualization (Power BI), Exploratory Data Analysis. Final Project: Maji Dogo Water Crisis Analysis.

Bachelor of Engineering (B.Eng.), Civil Engineering

2017 – 2023

Federal University of Technology - Minna

Coursework in Mathematics, Statistics, Numerical Analysis, and Python Programming.

Technical Stack & Domain Competencies

Core Engineering Matrix

A comprehensive overview of frameworks, languages, cryptographic tools, and predictive modeling libraries.

Data Science & Machine Learning

Scikit-learnPredictive ModelingAnomaly DetectionTime-Series & FFTK-Means ClusteringSupervised LearningStatistical Analysis

Data Analytics & Engineering

Python (Pandas, NumPy)Advanced SQL (PostgreSQL, MySQL)Exploratory Data Analysis (EDA)Excel (Pivot, VLOOKUP)Power BITableauStreamlit

Web3 Analytics & Tools

Flipside CryptoDune AnalyticsAlliumDefillama APIMEV Bot DetectionSolana / Base L2 AnalyticsBitquery GraphQL

AI & Full-Stack Development

PydanticAIGoogle Gemini SDKFastAPISupabase (RLS & Auth)Next.js 15 / ReactTailwind CSS