Daniel Akhabue

Senior AI Engineer · Senior Data Scientist · Kaggle Expert (#416 Global Rank)

I build intelligent systems that work in production: not just in notebooks. From multi-agent banking AI to computer vision security systems, I bridge cutting-edge AI with real enterprise impact.

I'm a Senior AI Engineer and Data Scientist based in Lagos, Nigeria, with 4+ years of experience building and deploying production-grade AI systems across banking, healthcare, government, and EdTech.

I specialize in the full AI lifecycle: from data engineering and model training to LLM orchestration, multi-agent systems, and cloud-native MLOps on AWS and Azure. I don't just build models; I build systems that enterprises trust.

• 🏆 Kaggle Expert: Global Rank #416
• 🥇 4x Hackathon Winner: including 1st Place at Zaka 2023 AI Challenge & Deep Learning IndabaX Nigeria
• 🎙️ International Speaker: PyCon Kenya 2025 · Datafest Africa 2024 · Black in Robotics 2025 (USA)
• 🏫 AI Faculty, Univaciti TeSA Program: training the next generation of production-ready AI engineers
• ☁️ Contributed to Qucoon's AWS AI Competency in Generative AI & Machine Learning Services
• 🌍 Organising Team Member & Faculty, AISOC (AI Summer of Code): the world's largest community for incubating AI talents
• ✍️ I write about AI strategy and engineering at Quantum Reports on Medium
• ⚡ Fun fact: I'm a certified Data Bender: I make data do things it didn't know it could

Problem: Enterprise digital banking overwhelmed by customer enquiries, manual transactions, and fragmented multilingual support
Solution: Pioneered a serverless, multi-agent conversational AI system on AWS Bedrock + Lambda with intelligent routing, stateful context management, and PCI-DSS compliant core banking integration
Impact: Scaled transactional banking by 70% · Cut customer enquiry workloads by 66% · Achieved >99% accuracy with sub-second latency
Stack: AWS Bedrock Lambda RAG Multi-Agent Systems Python

Problem: Traditional surveillance unable to process multi-modal, real-time inputs for comprehensive threat detection
Solution: Architected a CV pipeline fusing drone imagery, satellite feeds, thermal cameras, and ground sensors for 24/7 anomaly and intrusion detection with geospatial analysis
Impact: Enabled border security, infrastructure protection, and remote area surveillance at scale
Stack: PyTorch OpenCV Geospatial AI AWS Deep Learning

Problem: Manual EdTech content development was slow, expensive, and inconsistent across global cohorts
Solution: Built a proprietary AI model for automated course generation using Azure AI, integrated with Agentic student assistants for real-time learning support
Impact: Reduced content development cycles by 95% · Improved query resolution by 60% across 10K+ student interactions
Stack: Azure AI LangChain Agentic AI Python FastAPI

Problem: E-procurement customers faced slow, fragmented data access and poor support experience
Solution: Led design of an Agentic AI support system backed by a 500+ test case evaluation framework covering hallucination detection and tool-use accuracy
Impact: Accelerated data access and purchase information retrieval by 90%
Stack: Python LangChain SQL MLOps Drift Detection

🗂️ See all projects → bit.ly/dan-akhabue

current_focus = {
    "at_work":     ["Multi-Agent AI Systems", "IoT + Computer Vision Fusion", "AWS Bedrock"],
    "research":    ["LLM Fine-tuning", "RAG Optimization", "AI Evaluation Frameworks"],
    "teaching":    ["Production AI Engineering", "Cloud-Native MLOps", "Generative AI Systems"],
    "competing":   ["Kaggle", "Zindi Africa"],
    "north_star":  "Is the AI solving the problem it's expected to solve, under acceptable conditions?"
}

• 📄 My Last Umoja Hack Africa Experience as an Undergrad: The Good, the Bad, and the Ugly
• 📄 Driving Electoral Integrity with Data Visualization: How a Power BI Dashboard Can Revolutionize Your Election Process
• 📄 Is Hacking Together Your First Python Package Difficult?: A guide for data students
• 📄 Becoming a Contributor to the Open-Source Data Community
• 📄 Exploratory Data Analyses: NPFL's Teams and Standings: A novel data mining & EDA project

📚 Read more → Quantum Reports on Medium

Open to consulting engagements, speaking invitations, research collaborations, and conversations about hard AI problems.

📧 akdaniel0009@gmail.com