Ridwan Bello Salahuddeen

Highly skilled Machine Learning Engineer with over 10 years of experience in developing and deploying high-performance ML systems, specializing in Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) systems.

Available

LLM & RAG Specialist

Expert in Large Language Models, Retrieval-Augmented Generation, and Production AI Systems

LLM Development

Fine-tuning, optimization, and deployment of transformer models for production use cases

RAG Systems

Building scalable retrieval-augmented generation systems with vector databases

Production AI

Deploying and scaling AI systems with high throughput and low latency

About Me

Professional Summary

Results-driven Machine Learning Engineer with over 10 years of experience identifying opportunities for business growth and efficiency through advanced data analysis. Proven track record of leading the end-to-end technical execution of data science projects, from defining data acquisition and analytics strategy to deploying production models on cloud platforms (AWS, Azure, GCP).

Expertise in directing junior data scientists, managing stakeholder relationships, and presenting results to senior leadership. Extensive commercial experience in Python, Spark, and deep learning frameworks (PyTorch, TensorFlow) to build and scale ML solutions.

Machine Learning
Cloud Platforms
ML Infrastructure

Personal Information

Full Name

Ridwan Bello Salahuddeen

Location

Atlanta, Georgia, United States

Email

ridzy619@yahoo.com

Phone

+1 470-778-1174

Professional Experience

ML Engineer, Legal Document Processing

Apple (Contractor)
Feb 2024 – Present
  • LLM Optimization: Led the optimization of transformer models (BERT, RoBERTa) for high-throughput document processing, achieving 99% accuracy at 500+ pages/minute, while implementing efficient GPU resource management and monitoring systems.
  • RAG System Architecture: Engineered a novel hybrid chunking strategy combining small and large segments for improved RAG performance, reducing hallucination by 45% and influencing the team's approach to model optimization.
  • Vector Database Implementation: Architected and led the development of a scalable document embedding system using ChromaDB, processing and indexing millions of legal documents with semantic search capabilities.
  • LLM Fine-tuning: Implemented parameter-efficient fine-tuning techniques (LoRA, QLoRA) for domain-specific legal document understanding, reducing training costs by 60% while maintaining model performance.
  • Production Deployment: Designed and deployed production-ready LLM serving infrastructure with FastAPI and Triton Inference Server, handling 1000+ concurrent requests with sub-second latency.

Co-Founder & Technical Lead

Byte Analytics
August 2022 – Present
  • Led the design and development of system architecture for a scalable financial reconciliation system, directing a team of engineers.
  • Engineered and delivered regulatory International Financial Reporting Standards 9 (IFRS 9) compliance systems for two top banks in Nigeria, managing projects from conception to deployment and stakeholder acceptance.

Senior Machine Learning Engineer

BeeKeeperAI (Healthcare) & RigrAI (Video)
Jan 2023 – Feb 2024
  • LLM Fine-tuning: Directed the fine-tuning and deployment of a Llama model for forum summarization on darkweb data, utilizing weight orthogonalization and LoRA, and presented findings on model safety to stakeholders.
  • Model Serving Infrastructure: Engineered scalable model serving infrastructure with FastAPI, Docker, and Nvidia Triton Inference Server, implementing comprehensive monitoring and load balancing for efficient video processing.
  • Computer Vision & LLM Integration: Led the experimentation and training of indoor location (city, country and coordinates) detection models based on images, used for law enforcement and security purposes.
  • NLP Token Classification: Led the implementation and deployment of a BERT-based token classification model for PHI detection, achieving 97.3% F1 score.
  • Multi-modal AI Systems: Developed end-to-end pipelines combining computer vision models with LLM reasoning for enhanced security and surveillance applications.

Senior Data Science Engineer

Tribes.ai
Jan 2022 – Nov 2022
  • Led development of core Apache Airflow pipeline templates for auto-generating client-specific pipelines, reducing processing time by 45% and accelerating delivery of data products.
  • Optimized large-scale batch ETL using Spark (Dataproc), BigQuery, and Datastore for improved performance, showcasing experience with distributed data processing.

Enterprise Data Science Engineer

Sterling Bank PLC
Sep 2021 – Jan 2022
  • Led the end-to-end design, development, and deployment of fraud detection models, improving F1-score performance and presenting outcomes to business leaders to inform strategy.
  • Architected and optimized data processing workflows using Azure Synapse and Databricks for production deployment, demonstrating experience with cloud-based data processing.

Data Science Engineer

Access Bank PLC
Mar 2017 – Sep 2021
  • Developed and deployed a real-time transaction monitoring system using anomaly detection models and Spark Streaming, achieving a 46% reduction in the false-negative rate.
  • Drove initiatives for database performance optimization through efficient schema design and indexing on cloud infrastructure (AWS), reducing query latency by 30% and influencing data strategy.

Technical Skills

ML Frameworks

PyTorch 95%
TensorFlow 90%
JAX 85%
NLP (LLMs, Transformers) 92%
Keras 90%
scikit-learn 92%

ML Infrastructure

Model Serving (Triton, FastAPI) 88%
Distributed Training (NCCL) 85%
Pipeline Design 90%
Monitoring & Validation 87%

GPU Acceleration

CUDA 85%
CuDNN 82%
TensorRT 80%

Data Engineering

Apache Airflow 90%
Spark 88%
PostgreSQL 85%
Neo4j 80%

Cloud & DevOps

AWS 90%
Azure 85%
GCP 80%
Kubernetes 82%
Docker 88%
CI/CD 85%

Programming

Python 95%
SQL 90%
R 85%
Java 75%
Shell Script 80%

LLM & RAG Technologies

Large Language Models 95%
RAG Systems 92%
Transformer Models 90%
Fine-tuning (LoRA, QLoRA) 88%
Vector Databases 85%
Prompt Engineering 90%

Education

Master of Science in Machine Learning

Mohamed bin Zayed University of AI (MBZUAI)
2022 – 2024

Specialized in advanced machine learning techniques and applications.

Master of Science in Statistics and Computer Science

Georgia State University
2023 – 2025

Combined statistical methods with advanced computer science applications.

Bachelor of Technology in Electrical and Electronics Engineering

Ladoke Akintola University of Technology
2009 – 2014

Foundation in engineering principles with focus on electrical systems.

Certifications

  • Microsoft Certified: Azure Data Scientist, Data Engineer, and AI Engineer
  • Udacity: Deep Learning Nanodegree, Data Engineering Nanodegree
  • End-to-End ML with TensorFlow on GCP

Projects & Publications

Pothole Anomaly Detection

Research paper critically appraising approaches for real-time pothole detection in developing nations.

Read Paper →

Deep Learning Model for Potholes Data Augmentation

Exploratory analysis of modified deep learning models for potholes data augmentation at IEEE Conference.

Read Paper →

Hyperparameter Optimization Thesis

Master's Thesis on efficient hyperparameter optimization for large-scale ML systems.

View Thesis →

Hyperparameter Tuning Costs Research

Research on reducing hyperparameter tuning costs in ML, Vision and Language Model Training Pipelines.

Read Paper →

Get In Touch

Contact Information

Email

ridzy619@yahoo.com

Phone

+1 470-778-1174

Location

Atlanta, Georgia, United States

Connect With Me

Send Me a Message