Mohamed Muradh Maricair
London, United Kingdom | --|in/mohamedmuradh/
PROFESSIONAL SUMMARY
xperienced Data Scientist with over 4 yearsofexpertiseindevelopingAIsolutionstodrivedata-drivendecisions
E
and transform strategic capabilities. Proficient in SQL, Python, and employing machine learning libraries and
frameworks to meticulously manage and analyze large datasets across platforms. Skilled at navigating the
complexities of data collection, cleaning, and pre-processing, leading to the training and deploying of modelsthat
significantly impact revenue and key performance indicators (KPIs) growth.
SKILLS
●
●
●
●
anguage & Tools: Python, SQL, Tableau, Google Cloud,Azure, Docker, DevOps, Git, JIRA.
L
Technical Proficiencies: Data Pipeline, Data Management,Statistical Analysis, Software Engineering
Leadership Skills:Stakeholder management, Communication,Teamwork, Problem-solving, Agile.
ML, Statistics &LLM:RegressionAnalysis,PredictiveModeling,Bayesiananalysis,hypothesis&A/Btesting,
forecasting, time series analysis, TensorFlow, NLTK, Airflow, SciPy
WORK EXPERIENCE
ata Consultant
D
Jan 2023 - Present
Eternosoft
● Improved brand loyalty and customer engagement through predictive modeling and customer segmentation
techniques, increasing customer engagement by 30%.
● Implementedcustomizedmarketingstrategiesbasedondetailedcustomerbehavioranalysis,leadingtoa20%rise
in sales over six months and successfully countering prior sales declines.
ata Scientist
D
Sept 2022 - Dec 2022
University of Hertfordshire
● LedtheenhancementofanAIspeechtherapyapp,applyingdecisiontreesonunstructureddatatoboostprediction
accuracy to 83%, directly improving therapy outcomes for students with speech disorders.
● Managed an end-to-end initiative that developed a machine learning model with 88% accuracy, enabling early
detection and intervention for students, thereby personalizing and optimizing their learning paths.
● Enhanced user experience by designing an intuitive FAST-API interface and ensured application scalability by
implementing Kubernetes, contributing to a more accessible and reliable educational support tool.
achine Learning Engineer
M
Feb 2021 - Aug 2021
Sportifan
● SpearheadedinnovationatSportifan,anearly-stagestartup,leveragingMachineLearningtodelivermarket-ready
products like simulators and a Player Comparison system within the sports ecosystem.
● Developedafinancialpredictionsimulationtoolforsportsdirectorswith83%accuracy,utilizingPythonandSQL
for data analysis, directly contributing to strategic planning and financial management efficiencies.
● Effectively engaged with diverse stakeholders to meet objectives in the dynamic sports technology sector.
Data Analyst
Feb 2018 - Sept 2022
ccenture
A
● Deliveredover5000customizeddatarequeststoclientsutilizingadvancedSQLqueriestoderiveinsightsfromthe
customer relations database, identifying key correlations and anomalies to inform strategic decisions.
● Collaborated closely with Data Engineers across various departments to align data management and processing
techniques with industry best practices, resulting in a 15% improvement in targeted product recommendations.
● Designed, developed, and published 12 interactive Tableau dashboardsthatincreasedclientengagementby50%
and provided clear insights into data consumption per stack.
PROJECTS
tudent Performance Prediction|Python, scikit-learn,Random Forest, XGBoost, Flask, AWS, CI/CD
S
● Createdanend-to-endMLprojectaimedatpredictingstudentperformance,addressingkeyeducationalchallenges
by analyzing diverse factors including demographics, socioeconomic status, and academic records.
● Transformed the data by implementing MICE algorithm formissingdata,trainedwithCatboostregressormodel
which resulted in 84% accuracy.
ducators can now forecast student performance and identify students who need additional assistance using a
● E
predictive model encapsulated within a Flask web application.
AI content detector|NLP, PyTorch, transformer, BERT, RoBERTa, GCN, Word2Vec
● Conductedin-depthresearchonclassifyingAItext,leadingtothedevelopmentofthe'humanVSai'dataset,which
includes over 500 human and 1500 machine-generated texts from models like GPT-2, Neo and LSTM.
● Built a text classification model utilizing a Graph Convolutional Network (Text-GCN) with Word2Vec
embeddings,andfurtherexploredadvancedtechniquesbyimplementingBertGCNandRoBERTaGCNtoleverage
pre-trained transformer embeddings for enhanced performance.
● Analyzed the efficiency of GCN for text classification with large text data, proving 88% accuracy with
RoBERTaGCN over other models.
Favorita Grocery Sales Prediction |scikit-learn,pandas, numpy, seaborn, Decision tree, neural network
● Optimizedthelargedatasetwithover200,000differentproductsacrosssupermarketsforprocessingefficiencyby
implementing quantization techniques.
● Incorporated daily oil prices of Ecuador as a key feature in the Favorita Grocery Sales prediction model,
recognizing the country's economic sensitivity to oil price fluctuations.
● Utilized the PyTorch framework to train a neural network model, achieving a prediction accuracy of 78%.
Medical Chatbot |Langchain, huggingface, FaissDB,openAI
● Building a chatbot using LangChain framework and Llama 7-B model to enhance user engagement and
information accessibility.
● Build an RAG system on the vector embeddings to enhance contextual information retrieval.
● Employed Parameter-Efficient Fine-Tuning (PEFT) methodsontheHarvardn2c2datasetstorefinethechatbot's
responses, ensuring accuracy and faster retrieval in medical consultations.
POSITION OF RESPONSIBILITIES
tudent Representative, University of Liverpool
S
Oct2021- Aug 2022
● Involved in monthly SSLC meetings to discuss student issues and proposed solutions to the module and
programme coordinators. [Problem Solving]
● Published an enhancement report to the university based on the collective feedback of students and suggested
changes to the structure of the program. [Research, Planning]
Vice President, Rotaract Club of SVCE
Jun2015 - Apr 2016
● Demonstrated effective leadership by guiding a team of 100 students in planning, promoting, and executing
successful events. [Leadership]
● Engaged with management regularly regarding club development and upcoming events.[Communication]
EDUCATION
.Sc Data Science and AI
M
University of Liverpool
.TechChemical Engineering
B
Anna University
iverpool, UK
L
Sept2021- Dec 2022
hennai, India
C
July 2013 –May 2017
CERTIFICATES / ACTIVITIES
G
●
enerative AI with Large Language Models - DeepLearning.AI
● Deploying Machine Learning Models in Production – DeepLearning.AI
● Tableau Desktop specialist Certification – Tableau
ov 2023
N
May 2023
Aug 2020
RECOMMENDATION
● ‘ Muradh is a hard-working IT professional and expert Data Scientist, with high commitment to learning and
growth.HiscourseworkatCellStratAItrainingprogramwasimmaculateandhisAIprojectonFaceRecognition
intuitive.Muradhwilldowellduetohisdrivenattitude,collaborativemindsetandconstantdesiretolearn.Heis
friendly, humble and a team player.’
– Vivek Singhal, Co-Founder & Chief Data scientist, Cellstrat
