Ashik M Asharaf

Data Analytics Consultant | Brand Nerd

02/2025 – 03/2026 | Remote

• Social Media Analytics: Analysed cross-platform data (IG/TikTok/LI), identifying content pillars that boosted organic reach by 20% in 90 days.
• Automation: Built a Python/SQL pipeline to consolidate multi-channel data, reducing weekly reporting by 15 hours.
• Retention Strategy: Developed a Loyalty Dashboard to track engagement, identifying "at-risk" followers to maintain client retention.
• A/B Testing: Executed promotional campaign tests to optimise pricing and promotions strategies, protecting client margins.

Operations Analytics Manager | No3 Pubs UK Limited

03/2025 – 03/2026 | Manchester, UK

• Financial & Supply Chain Analytics: Optimised stock levels and reduced operational wastage by 20% through trend analysis of P&L statements.
• Business Intelligence: Developed a Customer Loyalty Dashboard to track repeat engagement patterns, translating metrics into growth strategies.
• Operational Leadership: Directed a team of 12 using performance mapping and data-driven KPIs to ensure high operational efficiency.

Corporate Trainer | CFG

05/2024 – 03/2025 | Manchester, UK

• Team Building: Recruited and trained a high-performing sales team; developed a data-backed sales pitch that increased lead conversion rates.
• KPI Management: Analysed team performance metrics to identify skill gaps, delivering targeted training that improved overall team efficiency.

Research Intern | RADD Solutions

09/2023 – 04/2024 | Hull, UK

• Developed an AI-powered framework for summarizing email content, leveraging supervised and unsupervised ML techniques.
• Designed and fine-tuned transformer models (T5, BART), achieving a 35% improvement in processing efficiency.
• Benchmarked model performance with precision metrics, achieving a ROUGE score of 46.3.
• Visualized insights using Power BI and Tableau, presenting results to stakeholders to drive strategic decisions.

MSc in Artificial Intelligence and Data Science

• Graduated with Distinction from the University of Hull.
• Specialized in Machine Learning, Data Analytics, Natural Language Processing (NLP), and Data Visualization.
• Completed a capstone project focused on AI-driven email content summarization, achieving a 35% improvement in summarization efficiency.
• Proficient in advanced tools and platforms like TensorFlow, PyTorch, Tableau, Power BI, and cloud environments (AWS, Microsoft Azure).
• Developed skills in statistical data analysis, optimization techniques, and neural networks.

BSc in Physics

• Graduated with First Class from Pondicherry University.
• Extensive experience in experimental physics, including laboratory-based research and practical experimentation.
• Hands-on knowledge in **Computational Physics**, simulating complex systems using numerical methods and Python programming.
• Deep understanding of **Laser Theory** and applications in spectroscopy and optics.
• Strong expertise in advanced mathematics, including calculus, linear algebra, and statistical methods applied to physics research.
• Collaborated on academic projects focused on energy systems and quantum mechanics, presenting findings in a university seminar series.

Email Content Extraction

Aim: To automate the summarization and extraction of key insights from unstructured email datasets to enhance processing efficiency.

Methodology: Leveraged transformer models such as T5, BART, and LongT5 for supervised learning and fine-tuned them using Python and TensorFlow. Built an unsupervised learning model for handling unlabeled email datasets. Employed data visualization tools like Tableau to present outcomes effectively.

Outcomes: Improved data processing efficiency by 35% and reduced computational resource requirements by 40%. Achieved ROUGE scores of up to 46.3, demonstrating high summarization quality.

Crime Hotspot Prediction

Aim: To predict crime hotspots using geospatial data, enabling better allocation of resources and crime prevention strategies.

Methodology: Utilized geospatial data and machine learning algorithms like XGBoost for hotspot prediction. Processed data using PostGIS for spatial analysis and visualized insights through Tableau dashboards.

Outcomes: Enhanced prediction accuracy by 20%, leading to improved resource allocation. Created visualizations that allowed stakeholders to identify and address high-risk areas effectively.

British Airways Customer Insights

Aim: To derive actionable insights from customer reviews and improve service quality and customer retention.

Methodology: Scraped and analyzed over 1,000 customer reviews using Python for sentiment analysis. Developed predictive models to assess customer behavior and utilized Power BI to build interactive dashboards for productivity tracking.

Outcomes: Improved customer retention by 15%. Delivered insights that enhanced customer satisfaction and streamlined decision-making processes through comprehensive data visualization.

Facial Emotion Recognition

Aim: To design a deep learning model capable of classifying emotions from facial images for applications in mental health and user behavior analysis.

Methodology: Built a hybrid CNN-LSTM model using Python, TensorFlow, and Keras to extract temporal and spatial features from image datasets. Enhanced the dataset with OpenCV for pre-processing and augmentation.

Outcomes: Achieved an accuracy of 85%, surpassing benchmarks in facial emotion classification. Delivered robust predictions suitable for real-world applications in mental health and human-computer interaction.

Breast Cancer Detection

Aim: To develop an image classification model for early detection of breast cancer from histopathological images.

Methodology: Utilized convolutional neural networks (CNNs) implemented in TensorFlow and Keras for image classification. Employed data augmentation techniques to improve model generalization. Preprocessed images with OpenCV and NumPy.

Outcomes: Achieved 92% accuracy in classification, supporting early diagnosis efforts. Reduced false negatives by 18%, providing reliable outcomes for clinical decision-making.

Financial Report Chatbot

Aim: To develop an AI chatbot capable of extracting and summarizing key insights from financial reports.

Methodology: Built a chatbot using Python and Flask, integrating GPT models for natural language understanding. Extracted data from 10-K and 10-Q reports using NLTK for parsing and analysis.

Outcomes: Streamlined financial data retrieval for decision-makers, reducing search time by 50%. Delivered concise insights, improving accessibility to critical financial metrics.

IBM HR Data Analytics

Aim: To analyze workforce data to predict employee attrition and optimize HR operations.

Methodology: Conducted data cleaning and preprocessing using Python and Pandas. Built predictive models with Scikit-Learn and visualized attrition risks using Power BI dashboards.

Outcomes: Reduced employee turnover by 12% by identifying high-risk attrition groups. Improved report generation efficiency by 30%, aiding HR in strategic planning.

Census Data Analysis

Aim: To analyze census data to provide insights into urban planning and resource optimization.

Methodology: Cleaned and preprocessed the dataset to handle missing values. Analyzed demographic trends, household occupancy, and commuting patterns. Visualized insights using Python libraries such as Matplotlib and Seaborn.

Outcomes: Provided actionable recommendations for housing optimization, skill development centers, and transport infrastructure improvements.

Accident Data Analysis

Aim: To analyze UK road accident data to classify accidents by severity and suggest preventive measures.

Methodology: Implemented clustering and predictive modeling techniques using Python, Scikit-Learn, and Tableau. Visualized accident hotspots and provided actionable insights.

Outcomes: Identified accident hotspots and proposed solutions to improve traffic safety and resource allocation efficiency.

Atmospheric Turbulence Models for Safer Aircraft Navigation

Aim: To explore atmospheric turbulence models that improve aircraft navigation safety.

Methodology: Conducted a literature review using scholarly sources, analyzing atmospheric conditions and their impact on aviation safety.

Findings: Identified key patterns in turbulence prediction techniques and explored the potential of machine learning and statistical forecasting models.

Real-Time Hate Speech Detection on Twitter

Aim: To develop a real-time hate speech classification model using machine learning and Apache Flink.

Methodology: Implemented a Twitter API-based data pipeline, analyzed existing hate speech detection models, and integrated a machine learning classifier.

Outcome: Successfully detected and classified hate speech tweets in real-time, enhancing automated content moderation.

Programming and Tools

• Python
• R
• SQL
• PySpark
• TensorFlow
• PyTorch
• Keras
• Scikit-Learn
• XGBoost
• LightGBM

Data Analytics and Visualization

• Power BI
• Tableau
• Excel (Advanced)
• Matplotlib
• Seaborn
• Plotly

Machine Learning and AI

• Supervised and Unsupervised Learning
• Natural Language Processing (NLP)
• Transformer Models (BERT, GPT, T5)
• Deep Learning (CNN, RNN, LSTM)
• Predictive Analytics

Data Engineering

• ETL Processes
• Data Quality Management
• Data Pipelines
• Data Warehousing
• APIs
• NoSQL Databases

Cloud Platforms

• Google Cloud Platform (BigQuery)
• Microsoft Azure
• AWS (S3, EC2)

Soft Skills

• Stakeholder Collaboration
• Documentation
• Teamwork
• Leadership
• Process Improvement
• Problem-Solving
• Adaptability
• Resilience
• Attention to Detail

IBM Data Engineering Professional Certificate

• Completed comprehensive training on data engineering processes, including ETL, data wrangling, and pipeline development.
• Acquired proficiency in data handling using Python, SQL, and NoSQL databases.
• Demonstrated expertise in data visualization using Tableau and Power BI.

Google Cloud Platform Big Data and Machine Learning

• Explored advanced machine learning techniques and their deployment using Google Cloud’s BigQuery and AI tools.
• Built scalable machine learning pipelines and big data solutions.
• Specialized in data processing and analysis on Google Cloud Platform.

GDPR Compliance Training

• Completed training on data privacy and GDPR regulations.
• Implemented compliance measures in data analysis and storage workflows.
• Ensured secure handling of sensitive data in all projects.

British Airways Data Science Simulation

• Analyzed over 1,000 customer reviews to generate actionable insights and improve retention by 15%.
• Developed Power BI dashboards for productivity tracking and sentiment analysis.
• Applied advanced data visualization techniques to present findings effectively.

Accenture Data Analytics and Visualization Simulation

• Cleaned and analyzed seven datasets to uncover trends and improve strategic decision-making.
• Built Power BI dashboards to visualize content trends and presented insights to stakeholders.
• Demonstrated expertise in data storytelling and actionable insights generation.

IELTS Academic Band 7

• Achieved Band 7, showcasing strong proficiency in English reading, writing, speaking, and listening.
• Applied effective communication skills in professional and academic environments.