JBU

AI Solutions Architect

Joseph Boi-Ukeme

Ph.D. researcher turned enterprise AI builder. I design production-grade systems — conversational analytics, autonomous agents, and ML pipelines — that turn complex data into decisions at scale.

Get in Touch LinkedIn GitHub Google Scholar
94.2%
Best Model F1
4.5M
Records / Day
200+
Platform Users
$4.2M
Projected Savings
50K+
Surveys Analyzed
About

Background

Researcher, engineer, and educator with a Ph.D. in Electrical & Computer Engineering. I specialize in bridging the gap between academic research and production systems — building AI solutions that work at enterprise scale.

Currently leading data science and AI initiatives for a federal government agency, advising on ML strategy, and collaborating with academic institutions on applied research. Previously a university lecturer in systems engineering, embedded systems, and data science, with published work in cyber-physical systems, fault detection, and network simulation.

Education

Ph.D. Electrical & Computer Engineering Carleton University, Ottawa — 2023 Thesis: “A Robust Data-Driven Method for the Design of Cyber-Physical Systems”
Grad. Certificate, BI System Infrastructure Algonquin College, Ottawa — 2020 CGPA: 3.93 / 4.0
M.Sc. Petroleum Engineering & Project Development IFP School / University of Port Harcourt — 2012 Best Graduating Student, Drilling Engineering
M.Eng. Electronics & Telecommunications University of Benin, Nigeria — 2011 Distinction
B.Eng. Electrical & Electronic Engineering University of Benin, Nigeria First Class Honours
Selected Work

Flagship Projects

Enterprise-scale AI with measurable business impact

Conversational BI Platform Generative AI

Conversational BI Platform

Enterprise analytics that translates plain English into SQL and interactive visualizations. Deployed to 200+ analysts across 6 departments, replacing multi-day report cycles with instant self-service insights.

  • 200+ active users
  • 6 departments
  • 70% faster insights
Databricks · GPT-4o · Dash · Plotly
CSFI Sentiment Analysis Machine Learning

CSFI Sentiment Analysis

Fine-tuned RoBERTa for multilingual government survey sentiment. Processes 50K+ citizen responses using BGE-M3 embeddings, powering real-time policy feedback loops.

  • 91% accuracy
  • 50K+ responses
PyTorch · HuggingFace · BGE-M3
Autonomous Agent Lab Agentic AI

Autonomous Agent Lab

Multi-agent orchestration using Model Context Protocol. Routes complex tasks across specialized agents with full observability, tracing, and error recovery.

  • 94% routing accuracy
OpenAI Agents SDK · MCP · asyncio
Portfolio

All Projects

GenAI

Procurement AI Assistant

Agentic SOW drafting with 87% first-pass compliance. Cuts manual review by 60%.

LangChain · NLP · python-docx GenAI

Document Classification Pipeline

3,400+ docs/month at 94.2% F1 using ensemble ML with LDA topic modeling.

Scikit-learn · XGBoost · spaCy GenAI

Groq Job Application Agent

Ultra-fast resume tailoring at ~12ms inference. 50+ applications per hour.

Groq LPU · Python ML

CSFI Predictive Analytics

Satisfaction prediction (R²=0.87) and department clustering with SHAP.

XGBoost · K-Means · SHAP ML

Event Stream Clustering

Real-time classification at 12K events/sec with sub-2s P95 latency.

Spark Streaming · K-Means · PCA ML

RAG Visualization Agent

PDF Q&A with auto-generated charts. 0.89 Recall@5.

ChromaDB · OpenAI · Streamlit Cloud

Enterprise Knowledge Hub

AI search across 2.4M+ docs. NDCG@10 improved 42%.

Vector Search · HNSW · FastAPI Cloud

Fleet Electrification Analytics

EV migration for 12K+ vehicles. $4.2M/yr projected savings.

Pandas · Geospatial · REST APIs Cloud

Network ETL Pipeline

4.5M records/day through Delta Lake. 98% success rate.

Databricks · Delta Lake · PySpark App

Vendor Connect Marketplace

Cross-platform marketplace with 68% Week-1 retention.

React Native · NestJS · PostgreSQL App

AI Report Generator

180+ reports/month in ~30s. Replaced 2–4 hour manual process.

Django · python-docx · GPT-4o App

PandasAI Project Reporting

NL queries on Jira data. 4 hours to 5 minutes reporting.

PandasAI · Jira API · Matplotlib
Experience

Career

2022 – Present

Technical Advisor, Data Science & AI

Federal Government Agency

Lead AI projects end-to-end. Advise on ML strategy, serve on technical committees for AI policy, collaborate with academia and AI vendors on applied research.

2020 – 2022

IT Analyst & A/Project Lead

Federal Government Agency

Progressed from BI to data science. Executed projects from planning through deployment. Led IT R&D strategic planning across government, academia, and industry.

2021 – 2023

Contract Instructor & Sessional Lecturer

Carleton University, Ottawa

Taught Operating Systems and Digital Systems at the undergraduate level. Designed curriculum, delivered lectures, supervised student projects.

2017 – 2022

Graduate Teaching & Research Assistant

Carleton University, Ottawa

Supported 13+ courses across real-time systems, OOP, operating systems, and data management. Supervised labs and advised students.

2012 – 2016

Lecturer, Electrical & Computer Engineering

University of Benin, Nigeria

Taught 10+ courses. Developed new curricula, supervised 7 capstone projects, served on the project defence committee.

Research

Publications & Service

Selected Publications

Dissertation

Boi-Ukeme, J. (2022). A Robust Discrete Event Method for the Design of Cyber-Physical Systems. Carleton University.

Journal

Rajus, V.S., Boi-Ukeme, J., et al. (2022). Measured Data Reliability for Building Performance. IEEE I&M Magazine, 25(1), 55–61.

Conference

Jamal, M., Boi-Ukeme, J. & Wainer, G. (2022). CAN DEVS for Independent Node Testing. ANNSIM, 790–801. IEEE.

Conference

Boi-Ukeme, J. & Wainer, G. (2021). Data-Driven Fault Detection in Buildings. Winter Simulation Conf., 1–12. IEEE.

Conference

Boi-Ukeme, J., Ruiz-Martin, C. & Wainer, G. (2020). Real-Time CPS Fault Detection in DEVS. IEEE EBCCSP.

Academic Service

Program Committees
  • ANNSIM — CNS Track 2021–2026
  • Winter Simulation Conf. — DSS Track 2020–2023
  • IEEE 5G and Beyond Working Group 2019–2021
Peer Review
  • IEEE Trans. on Industrial Informatics
  • ANNSIM · WinterSim · SpringSim · ACM/IEEE PADS
Expertise

Technical Stack

AI & Machine Learning

GPT-4oGroq LPURoBERTaXGBoostPyTorchScikit-learnSHAP

Data Engineering

DatabricksDelta LakeSparkPandasChromaDBFAISS

Frameworks & Orchestration

LangChainOpenAI Agents SDKMCPRAGHuggingFace

Development & Cloud

FastAPIDjangoNestJSReact NativeDashStreamlitAzureDockerMLflow

Let’s Build Something

Open to consulting, collaboration, and opportunities in enterprise AI.

joejazo@gmail.com LinkedIn GitHub Google Scholar