Silbaway Ra-is Osumanu
Machine Learning Engineer & NLP Specialist
Transforming complex data challenges into innovative AI solutions
Professional Summary
Result-driven and highly skilled Machine Learning Engineer with a specialization in Natural Language Processing (NLP). Possessing a solid foundation in mathematics, statistics, and computer science, I excel in designing and implementing innovative machine learning solutions to tackle complex problems in language understanding and generation.
With hands-on experience in developing and deploying state-of-the-art NLP models using frameworks like TensorFlow, PyTorch, SpaCy, and AWS SageMaker, I have a proven track record of delivering high-quality, scalable solutions that drive business value.
Technical Expertise
Machine Learning & AI
- TensorFlow & PyTorch
- Deep Learning & Neural Networks
- Machine Learning Algorithms
- Computer Vision
NLP & Chatbots
- Natural Language Processing
- Transformers & LLMs
- Rasa & Dialogflow
- SpaCy & NLTK
Development
- PHP & Laravel
- Python & FastAPI
- RESTful APIs & AWS
- Docker & Kubernetes
Certifications
Databricks Lakehouse Fundamentals
Databricks - February 8, 2024
Generative AI Fundamentals
Databricks - January 22, 2024
Open Source Models
Contributing to the AI community through open-source models on Hugging Face
Medical Generation Model
Fine-tuned T5 model for medical diagnosis
A specialized model designed to generate medical diagnoses and treatment recommendations. Trained on clinical scenarios to provide accurate and contextually relevant medical outputs.
Lyrics Generation Model
Fine-tuned GPT-2 for Ghanaian lyrics
A creative AI model trained on Ghanaian artist lyrics, capable of generating coherent and contextually relevant song lyrics based on input prompts.
Medical Generation Model with COT
Advanced medical diagnosis model with step-by-step reasoning
An enhanced version of the medical generation model that incorporates Chain of Thought reasoning for improved diagnostic accuracy. Fine-tuned on clinical datasets to provide detailed, logical medical assessments and treatment recommendations for healthcare professionals.
Professional Experience
Senior Software Engineer
Renobytes – Accra, Ghana (Remote)
June 2023 - Present
- Developed scalable applications using agile methodologies for timely project delivery
- Architected and implemented robust USSD applications and backend integrations using PHP/Laravel
- Developed custom sentiment analysis models tailored to specific industries
- Led migration efforts to cloud-based infrastructure for NLP applications
Senior Application Manager
Halges Financial Technology – Accra, Ghana
October 2017 - November 2023
- Built and maintained multiple USSD applications for financial services using PHP/Laravel
- Implemented advanced machine learning techniques for better understanding of complex language patterns
- Developed RESTful APIs and backend integrations for mobile money services
- Mentored junior developers in PHP/Laravel best practices and USSD application development
Teaching Assistant
University of Ghana, Computer Engineering Department
September 2016 - 2018
- Assisted professors in course management and material coordination
- Provided tutoring support and reviewed student assignments
- Facilitated a positive learning environment for computer engineering students
Education
Bachelor of Science: Computer Engineering
University of Ghana - Accra, Ghana
October 2012 - June 2016
Key Accomplishments
- 2019/2020 MTN MoMo and Ericsson Developer Hackathon Challenge Second Place Winner
- Successfully developed and deployed over 10 USSD applications using PHP/Laravel, serving thousands of users in Ghana's financial sector
- Reduced customer churn by 10% through implementation of machine learning models and robust backend systems
- Led the development of critical payment integration systems using Laravel, connecting multiple financial institutions
Featured Projects
MobiFund
View ProjectCrowdfunding Application
A prototype crowdfunding application enabling users to upload projects and receive contributions from supporters. Contributors can vote on fund disbursements requested by project creators. Features sentiment analysis on comments using machine learning for community feedback insights.
Recognized as a notable project under the MTN Platform Accelerator program in 2024
Blood Cell Classification
View on GitHubDeep Learning for Medical Imaging
Developed a machine learning pipeline to classify blood cell images into different subtypes using various approaches, including CNN, EfficientNet, and Vision Transformer (ViT). Achieved highest accuracy using the Vision Transformer model on a dataset of 12,500 images.
Customer Forecasting Dashboard
View on GitHubInteractive Data Visualization & Forecasting
A sophisticated dashboard application for customer behavior analysis and forecasting. Built with Facebook Prophet for accurate predictions and Plotly Dash for interactive visualizations. Features include historical trend analysis, pattern detection, and customizable forecast periods.
Learn For Me
Chatbot for Document Retrieval and Summarization
Developed an intelligent chatbot that allows users to upload documents, which are then converted into embeddings and stored in a vector database. Implemented a similarity search mechanism to retrieve relevant content based on user queries.
Bufei
Virtual Assistant Food Ordering Bot
Developed a virtual assistant food ordering bot for WhatsApp and Telegram, enabling user registration, login, and location sharing to find nearby restaurants. Built with PHP, Laravel framework, and JavaScript, hosted on AWS.
Sales and Customer Churn Prediction
Machine Learning Project
Developed and deployed a machine learning model achieving 95% accuracy for sales prediction with recurrent neural networks and 87% for churn prediction with XGBoost and Random Forest. Improved sales by 15% and reduced churn by 10%.
Plant Disease Prediction
Image Classification using Deep Learning
Developed an image classification model to predict plant diseases using a fine-tuned pre-trained ResNet architecture. The system accurately identifies various plant diseases from uploaded images, aiding in early detection and effective management of plant health.
Web Intelligence Summarizer
AI-Powered Web Research Assistant
Developed an intelligent web research system that automatically gathers and analyzes information about people from across the internet. When users input a name, the system leverages search engine APIs to fetch relevant content and posts, processes the data using OpenAI's language models, and generates comprehensive, categorized summaries of the person's online presence and activities.
Medical Generation Model
View on HuggingFaceChain of Thought Medical Diagnosis Model
Developed a fine-tuned medical diagnosis model incorporating Complex Chain of Thought (CoT) reasoning. The model enhances logical and step-by-step reasoning in clinical scenarios, providing contextually aware and clinically relevant responses for medical professionals and AI-assisted healthcare solutions.