Tarique Aziz

Responsible for relaying requirements to vendors for client requests.
Creating framework for testing rule engines like IBM ODM and SPSS for Suspicious Monetary transaction using python, flask, db2.
Integrated the application with ODM and SPSS through IIB using REST API.
Leading a team for the development of the portal for the Saudi Arabian Financial Investigation Unit to report suspicious financial transactions using Spring Java 8.0.
Designed a database and implemented a portal for fintech transactions.

Responsible for relaying requirements to vendors for client requests.
Creating framework for testing rule engines like IBM ODM and SPSS for Suspicious Monetary transaction using python, flask, db2.
Integrated the application with ODM and SPSS through IIB using REST API.
Leading a team for the development of the portal for the Saudi Arabian Financial Investigation Unit to report suspicious financial transactions using Spring Java 8.0.
Designed a database and implemented a portal for fintech transactions.

Created python analytics library for geospatial data analytics on aws
Designed a database using AWS Athena, AWS glue for querying geospatial data cost effectively from S3 buckets written by kinesis streams and created reports for clients to inform their marketing strategy. Achieved cost efficiency > 5 times.
Implemented custom clustering algorithm based on DBSCAN to create location based marketing analytics report
Implemented a high performance K-means clustering algorithm in C++, cuda(RTX GPU) and created a python library of the same using cython bindings.
Achieved a performance boost of 5 times compared to sklearn.
Implemented early surgical indicator prediction model using random forest and achieved AUCROC of .9
Utilized 1.5 years of medical history data to extract features involving medication administered, lab tests, visits to hospital.
Used Rapidminer, Python to clean the data and transform it into the format suitable for random forest.
Trained various models like svm, logistic regression and random forest and chose random forest model due to higher AUC ROC value.
Implemented high performance similarity detector for images using faiss library to recommend fashion products with response time < 5ms.
Used cv library to normalize images and convert them into HSV format.
Vectorised the image and indexed in faiss for measuring similar images based on cosine similarity.
Created a feed forward neural network classifier on Azure using pyspark and spark ml to classify products described in multiple natural languages with 99% accuracy.
Extracted data using elastic search from Azure for millions of correctly classified product samples and loaded into spark for cleaning and preprocessing.
Converted the natural language to TF-IDF vector and use sparkMLP to train a feed forward neural network by loading data in batches from spark

Implemented medical data harmonization project using SNOMED-CT ontology.
Exposed the services of the product using restful API in Spring Java.
Improved the recall of the search engine by 50%.
Integrated with allegrograph and created cron jobs to update data daily into postgresql.
Worked on backend of projects like KRSTE, Padipedia on Spring Framework
Implemented restful api to deliver data for visualization.
Implemented high performance graph analyzer software in CUDA, C++ for TESLA GPUS to analyze graphs with millions of entities and billions of connections.
Created scheduler to load data in GPU and calculated local closeness centrality, betweenness centrality, neighbors size etc.
Optimized legacy code to get performance boost >100 times.
Used in researcher discovery, anti terrorism, anti money laundering projects
Led development of accounting application in ruby on rails for automation of financial reports and supported marketing team with optimized SQL queries.
Integrated application with redis nosql database to get real time updates on rides.
Created cron jobs to updated databases daily with updated information.
Published a patent for semantic matching of textual data.
Invented a method for fuzzy matching of semantic graphs of documents using AI optimization techniques like Simulated Annealing and Genetic Algorithms.
Created machine learning and deep learning model for Dengue prediction in K. Lumpur
Collected weather data like rainfall, temperature from nasa and local government bodies and made them of uniform interval using cubic spline interpolation.
Used sklearn for multivariate polynomial regression and tensorflow for RNNs to create predictive models.

Created python analytics library for geospatial data analytics on aws.
Implemented custom clustering algorithm based on DBSCAN to create location based marketing analytics report.
Implemented a high performance K-means clustering algorithm in C++, cuda(RTX GPU) and created a python library of the same using cython bindings.
Implemented early surgical indicator prediction model using random forest and achieved AUCROC of .9.
Implemented high performance similarity detector for images using faiss library to recommend fashion products with response time < 5ms.
Created a feed forward neural network classifier on Azure using pyspark and spark ml to classify products described in multiple natural languages with 99% accuracy.

Implemented medical data harmonization project using SNOMED-CT ontology.
Worked on backend of projects like KRSTE, Padipedia on Spring Framework.
Implemented high performance graph analyzer software in CUDA, C++ for TESLA GPUS.
Led development of accounting application in ruby on rails for automation of financial reports and supported marketing team with optimized SQL queries.
Published a patent for semantic matching of textual data.
Created machine learning and deep learning model for Dengue prediction in K. Lumpur.