Experienced Data Scientist working in analytics consulting.
Proficiencies:
>Data analysis and KPI identification > Data Visualization > Machine Learning > Deep learning
> Conversational AI > Topic Modelling > Information Extraction > REST API
> Azure > Python > Numpy, Pandas, Matplotlib, Seaborn, GGplot > Keras
> KNIME > SpaCy > RASA > DataRobot
Add work experience to your profile. (optional)
Developed and implemented a monitoring system for identification of extremely high gas flares emitted by
upstream/downstream energy facilities across multiple geographies serving a leading Consulting firm in Oil & Gas domain
.Performed analysis and identified countries with facilities displaying patterns of persistent high flares.
• Created Pareto analysis charts for identification and segmentation of countries and facilities emitting most flares.
• Worked on a project to develop a comprehensive model for detection of correct pose in people involved in manual handling of
goods.Developed a method for the human pose estimation based on Deep Neural Networks by detection & localization of joints.
OCR – INVOICE LINES EXTRACTION
• Designed, developed and implemented an OCR service for a UAE based Retail major assisting in extraction of line items and
specific terms from invoices. Built and trained a pipeline of predictive models on sample dataset and achieved an F1 score of
91%. Utilizing a combination of Deep Learning, Image processing based methods for extraction of line items.
• Performed testing of the framework on multiple invoice types and assessed the model performance using key metrics. Created a
dashboard for monitoring of the model performance metrics for new batches.
COUNTERFEIT DRUG DETECTION
• Worked as analytics consultant for UK based Fortune 500 firm. Helped the client develop a process in accurately identifying
counterfeit drugs. Preparation of reference data pertaining to each drug category. Developed an alert system by building
predictive models to identify & segment counterfeit samples. Trained the predictive models on historical data and achieved an
F1 score of 98%. Implemented reporting the results of batches to the client.
CUSTOMER RELATIONSHIP PREDICTION
• Actively involved in development of a customer relationship management dashboard for a Middle Eastern Telecom leader to
improve customer engagement and retention.Performed analysis of the users data and identified relevant KPIs. Developed
predictive models to assess customer churn and determined the key factors leading to a user exiting the service.Helped the
client achieve significantly improved business outcomes
Add work certification to your profile. (optional)
Certified Datarobot platform – Data Science Associate
Certified KNIME Data Science developer
Add work education to your profile. (optional)
We will review the reports from both freelancer and employer to give the best decision. It will take 3-5 business days for reviewing after receiving two reports.