***** **** ***, ****** ****

Baltimore, *****

240-***-****

ad2167@r.postjobfree.com

Mohamed Alpha Kamara

Machine learning engineer with hands-on experience designing, developing, and deploying large-scale machine learning systems. Skilled in Python, Spark and optimizing systems for production environments. Strong foundation in statistical modeling, predictive modeling, and data mining. Proficient in developing and deploying large-scale distributed machine learning systems. Skilled in applying predictive modeling, deep learning, and statistical analysis to drive impactful business solutions. EXPERIENCE

AFCOM SL, Sierra Leone — Data Scientist

July 2017 - August 2020

● Apply statistical techniques and machine learning algorithms to analyze and interpret complex datasets.

● Collaborate with cross-functional teams to understand business problems and define data-driven solutions.

● Develop and deploy machine learning models for various applications such as recommendation systems, fraud detection, or demand forecasting.

● Create interactive dashboards and visualizations to enable self- service data exploration and analysis. • Ensure data quality, integrity, and security throughout the data lifecycle.

● Built regression models to forecast product demand, enabling a 10% reduction in inventory costs.

● Engineered an anomaly detection system for fraud analysis using isolation forest and PCA, improving fraud detection by 20%.

● Developed interactive Tableau dashboards to visualize customer segmentation insights, aiding strategic marketing campaigns.

● Performed A/B testing to evaluate effectiveness of recommendation algorithms. Optimized models to improve click- through rate by 18%.

● Cleaned data pipelines with PySpark, ensuring quality, security, and privacy for downstream analytics.

Clarus Enterprises, Remote — Machine Learning Engineer January 2022 – January 2023

● Worked with a cross-functional agile team on the design, development and production deployment of real-time retail product recommendation engines using PySpark and AWS SageMaker.

● Implemented LSTM-based deep learning models leveraging Spark and TensorFlow to generate product recommendations with improved click-through rates.

● - Developed and deployed deep learning recommendation systems handling over 50 million daily requests with 99.95% uptime

● - Built NLP semantic search engine to improve search relevancy by 15%; implemented text embeddings and fine-tuned BERT models

● - Led migration of machine learning services to Kubeflow, establishing CI/CD pipelines for model monitoring and governance

● - Communicated complex model explain ability and ethical AI topics to non- technical executives through presentations and reports

● Published TensorFlow recommendation model architecture whitepaper explaining innovations to industry peers.

● Created an ML content repository and guided 20+ engineers through hands-on labs on topics like transfer learning.

● Worked cross-functionally to produce computer vision models for defect detection, reducing false positives by 10%

● Continuously improved models through retraining on updated data, tracking key accuracy and fairness metrics.

● Optimized model inference latency from 600ms to 300ms and reduced overall memory footprint by 40% to enable cost-efficient scaling to production workload.

Informatics Consultancy, Freetown — Senior Machine Learning Engineer (Remote) January 2023 - Present

● Engineered real-time product recommendation system on Spark and Sage Maker, improving click-through rate by 15%.

● Reduced fraud by detecting anomalies in large-scale transaction data using isolation forests and neural networks.

● Presented data insights and scale-out proposals to executives, securing

$500K in additional funding for ML teams.

● Built deep learning models using TensorFlow and PyTorch to classify images and detect credit card fraud.

● Operationalized NLP classifiers on GCP to improve search relevance by 12%.

● Communicated regularly with stakeholders via dashboards and reports.

● Implemented deep learning models using TensorFlow for [application, e.g., image classification].

● Conducted A/B testing and optimized models to improve [specific metric, e.g., accuracy].

● Collaborated with data scientists and engineers to operationalize deep learning models for computer vision and NLP use cases on Google Cloud Platform leveraging services like Big Query, Dataflow and AI Platform

● Created interactive Tableau dashboards to communicate model quality metrics and business impact to senior stakeholders across the organization.

● Built anomaly detection algorithms for credit card fraud analysis achieving 99% precision. Published models via REST API endpoints secured using Auth0 for client usage.

EDUCATION

University of Maryland, Collegepark — MPS of Machine Learning August 2022 - December 2023, United States of America University of Reading, Reading — MSC Data Science & Computing September 2020 - December 2021, United Kingdom

University of Sierra Leone, Freetown — BSC Math’s & Statistics September 2014 - August 2019, Freetown

TECHNICAL SKILLS

● Programming Languages: Python, MATLAB, SQL, R, C, C++, Fortran, Bash, STATA

● Machine Learning: Supervised and Unsupervised Learning, Deep Learning, Natural Language Processing (NLP), Computer Vision, anomaly Detection

● Data Assimilation: Kalman filters, particle filters, optimal interpolation

● Cloud Infrastructures: AWS (EC2, S3, SageMaker, EMR), Azure(VM,Blob Storage, ML studio, HDinsight), GCP

services(Compute Engine, AI platform, BigQuery)

● Big Data: Spark, MPI/Slurm, Airflow, Kubernetes

● Data Analysis: Pandas, NumPy, Matplotlib, Seaborn, SPSS, STATA, ODK

● Machine Learning Frameworks: TensorFlow, PyTorch, scikit-learn

● Tools & Libraries: Jupyter Notebook, Git, Docker

● Database Management: SQL, MongoDB

● Data Visualization: Tableau, Power BI, Looker, D3.JS

● Mathematics: Calculus, Linear Algebra, Time Series Forecasting Projects

● Image Classification with Convolutional Neural Networks (CNNs)

Built a CNN model using TensorFlow to classify images in the CIFAR-10 Dataset.

Achieved an accuracy of 92%.

Implemented data augmentation techniques to improve model generalization.

● MTG Net Cards

Used a LSTM Model to predict the prices of a certain card.

Developed a Python-based NLP model using [Tensorflow and pytouch].

● Credit Card Fraud Detection

Built anomaly detection models achieving 99% precision using Random Forest, XGBoost and isolation forests.

● Forecasting Used Car Prices

Utilize cleaned Kaggle used car dataset.

Establish linear regression baseline.

Build and tune deep feedforward networks in TensorFlow/Keras

Leverage CNNs to identify local patterns.

Use regularization, ensemble techniques to improve accuracy.

Optimize MSE metric, target 10% improvement over baseline.

● Forecasting of Real-estate Dataset

Using Advance Machine Learning Models to show patterns and prediction in the United States Real estate Dataset from the realtor.com website

CERTIFICATIONS

● Machine Learning Certification

Simpli-Learn 2021

LANGUAGES

● English (Fluent)

Contact this candidate