Cliff Mbah

Developed and optimized end-to-end machine learning pipelines
Developed the Machine Learning Days to Depletion Algorithm (MLDDA) for Lexmark toner replacement mechanism with an accuracy of 95%. Model outperformed the previous non-ML system by 4%
Utilized algorithms such as XGBoost, LightGBM, decision trees, SVM, and deep learning to develop and improve a variety of
predictive models to optimize the Lexmark-customer toner delivery processes, ensuring better customer experience
Performed advanced hyperparameter tuning ( hyperopt , GridSearchCV and RandomSearchCV) to improve performance of existing
and new models for subscription devices
Develop and manage models in cloud platforms (azure databricks)
Track and log experiments, parameters, metrics, and artifacts with mlflow
Perform advanced exploratory data analysis (EDA), feature engineering and data preprocessing to transformed data into a suitable
form for model development
Leveraged evidently AI to monitor models in production by running drift reports to analyze input, performance, and target drifts
between reference (data used for training) and current data (data in production and wasn't used for training). Model is retrained
depending on drift values
Utilized programming languages such as Python, Pyspark, and SQL to build ML models use in driving key business decisions.
Utilized different visualization tools such as Power BI (Tableau) and Python visualization libraries (seaborn, matplotlib, and plotly) to
develop plots used in finding patterns, correlations, and trends in our data and present findings to stakeholders.
Developed an event anomaly detection service to identify anomalies using a time-aggregated data from a population of devices.
Work with stakeholders to develop quarterly roadmaps based on impact, effort, and test coordination.
Planned projects to be accomplished over a program increment by setting clear objectives and key results (OKRs).
Collaborated with cross-functional teams to develop data-driven solutions to business challenges
Provide comprehensive analysis and recommended solutions to address complex business problems using data from internal and
external sources and applied advanced analytical methods to assess factors impacting growth and profitability across product and
service offerings
Actively involve in mentoring junior data scientist
Write and present technical reports to both technical and non-technical stakeholders

Developed and optimized end-to-end machine learning pipelines
Developed the Machine Learning Days to Depletion Algorithm (MLDDA) for Lexmark toner replacement mechanism with an accuracy of 95%
Model outperformed the previous non-ML system by 4%
Utilized algorithms such as XGBoost, LightGBM, decision trees, SVM, and deep learning to develop and improve a variety of predictive models to optimize the Lexmark-customer toner delivery processes, ensuring better customer experience
Performed advanced hyperparameter tuning (hyperopt, GridSearchCV and RandomSearchCV) to improve performance of existing and new models for subscription devices
Develop and manage models in cloud platforms (azure databricks)
Track and log experiments, parameters, metrics, and artifacts with mlflow
Perform advanced exploratory data analysis (EDA), feature engineering and data preprocessing to transformed data into a suitable form for model development
Leveraged evidently AI to monitor models in production by running drift reports to analyze input, performance, and target drifts between reference (data used for training) and current data (data in production and wasn't used for training)
Model is retrained depending on drift values
Utilized programming languages such as Python, Pyspark, and SQL to build ML models use in driving key business decisions
Utilized different visualization tools such as Power BI (Tableau) and Python visualization libraries (seaborn, matplotlib, and plotly) to develop plots used in finding patterns, correlations, and trends in our data and present findings to stakeholders
Developed an event anomaly detection service to identify anomalies using a time-aggregated data from a population of devices
Work with stakeholders to develop quarterly roadmaps based on impact, effort, and test coordination
Planned projects to be accomplished over a program increment by setting clear objectives and key results (OKRs)
Collaborated with cross-functional teams to develop data-driven solutions to business challenges
Provide comprehensive analysis and recommended solutions to address complex business problems using data from internal and external sources and applied advanced analytical methods to assess factors impacting growth and profitability across product and service offerings
Actively involve in mentoring junior data scientist
Write and present technical reports to both technical and non-technical stakeholders

Utilized predictive analytics to develop a customer churn prediction system to optimize customer retention rates
Developed and managed end-to-end ML pipelines (from ingestion and integration to deployment)
Leveraged ML algorithms such as xgboost, random forest, deep learning, and others to build advanced models used in solving business problems and optimizing revenue
Monitor models in production for performance drift
Performed extensive exploratory data analysis, feature engineering, data preprocessing and other processes to ensure data is transformed into form suitable for model training
Identified valuable data sources and automate collection processes
Developed best-suited data models and algorithms to drive business decision
Enable smarter business processes by using analytics for meaningful insights
Utilized unsupervised ML techniques (KMean clustering) to develop customer segmentation models used to deliver more personalized and effective marketing campaigns, leading to increased customer retention and revenue
Collaborated with multi-disciplinary product development teams to identify performance improvement opportunities and integrate trained models
Supported research and development of improvements in analytic techniques and capabilities
Worked with a dynamic data science team as well as independently to execute data science projects, perform analysis, and present results and/or solutions to stakeholders
Coached and mentored junior data scientists on data modeling techniques

Leveraged ML algorithms to developed an insurance fraud detection model use in predicting fraudulent claims
Optimized data pipelines to improve data quality and accuracy
Developed algorithms to detect anomalies and uncover hidden patterns in customer data
Combined analytical applications with a continuous stream of real-time data to create detailed and personalized assessments of risk
Utilized visualization tools such as power BI and python libraries (plotly, seaborn, and matplotlib) to uncover different trends and patterns in data
Tuned models using different ML techniques to improve performance
Developed and maintained data models to support key business decisions
Created customized applications to make critical predictions, automate reasoning and decisions and calculate optimization algorithms
Used statistical tools to interpret data sets, paying particular attention to trends and patterns that could be valuable for diagnostic and predictive analytics efforts
Assessed effectiveness and accuracy of new data sources and data gathering techniques
Led discussions and assess the feasibility of AI/ML solutions for business processes and outcomes
Mentored several Jr. Data Scientists on the team