Muhammad Naeem

M.Naeem Sr. Data Scientist ..... About Address Multan, Pakistan Tel - Mail naeem.lyon@ gmail.com Web & Git linkedin github guru Education PhD InfoMathematics 2017 France PhD Machine Learn 2013 Pakistan MS Computer Sci. 2010 Pakistan BEng. Metallurgy 1997 Pakistan Deep Learning Transformer RNN,LSTM CNN PyTorch keras TensorFlow h2o OpenCV2 OCR Tesseract Textract pdf / images: MP HTR 20+ years of experience with 7 years of experience as Data Scientist using ML Algorithms, NLP & Computer Vision. Experience 09/22 - 11/22 Machine Learning Engineer (Remote) Freelance Work @ guru.com Design and development of a self-learning regression model to correct the prediction of bit pressure from the sensor installed in drilling. Regression Modeling • Identify and correlate parameters from IoT sensors to make unified data for the regression model • Identify changes and patterns indicating the error in the prediction of bit pressure in drilling • Compared state-of-the-art regression models and model optimization • Feedback mechanism which can re-train the model to learn from the previous mistakes • Tech: TensorFlow, XGBoost, Random Forest, Arima, Pandas 01/19 - 08/22 Sr. Data Scientist (Remote) Friendly Health Tech. USA Design, develop, deploy and maintain an AI-based document processing workflow automation system. Captured and analyzed historical policy data from noisy scanned docs in health, and medical form processing for insurance companies OCR ML Modeling • OCR hand-written and machine-printed text using ML techniques • Supervised the process of annotation of scanned data • Compared state-of-the-art HTR and OCR models. Fine-tuned the best model • Deployed models at AWS Sagemaker. Later moved to ECS and finally Kubernetes. • Tech: Deep Learning models, TensorFlow, Tesseract, AWS Textract, AWS sagemaker / ecs deployment, OpenCV2, Docker, Flask, Git. Layout ML Model Develop a system to identify questions and answers from different types of claim forms. • Developed an indigenous rule-based system to identify questions and their corresponding answers from free OCR text • Fine-tune the ML Layout Model V2 and then V3 upon a large number of annotated forms • Tech: Unified pre-training for language understanding (NLU) and generation (NLG) Rule Based Models Develop a recommendation system for insurance companies to help in Adjudication or Claim rejection/acceptance • Developed a set of hierarchical rules to convert unstructured text from ML models to populate the schema-based knowledge • Devised a set of rules to fetch necessary information from table data into schemas • Devised a set of rules to turn the schema text into actionable rules. • Tech: NLP, Spacy, Named Entity Recognition, regex, AWS Comprehend Medical, Docker, Flask, Git. Big Data Experience Hadoop mapReduce Spark noSql 05/17-11/18 Data Scientist (Urban Mobility) Deployment Docker Github AWS: Sagemaker ECS IFSTTAR/VEDECOM Versailles France • • • • GIS data analyst for daily vehicular urban mobility Development of business logic for human mobility analyzer Analysis of geometry-oriented delays in public transport Modeling the links b/w territorial attractors, and infrastructures connecting these attractors • Simulation of estimated displacements around territories of interest Tech: LSTM, open street map, google apis, leaflet. 09/15-04/17 Research Engineer NLP Sent.Analys. Sumarization NER SpaCy coreNLP NLTK AWS: DISP Lab. Lyon France • Development of front end & backend business process services for enterprise collaboration support & decision-making system • The system is used to analyze and identify a set of opportunities to boost industrial manufacturing by means of data analytics enablers such as Timing enablers, Costing, Resource, Risk, Quality Control, Product Specification (Functional, Technical, General), and Process across enterprises, Raw Material, Product assembly and handling, Production capacity, Customer Engagement & Customer Lifelong Value Tech: deep learning, machine learning, mongoDB, Docker. comprehend topicmodeling Databases MySQL Oracle MongoDB Microsoft: SQL Serv. Access/VBA Programming Python R Java C# C/C++ VB.Net Matlab 09/13-08/15 PhD Researcher (Enterprise Collaboration) ULL2 Lyon, France • Translation of structured and semi-structured data into value • Concluded state-of-the-art into a holistic & dynamic framework comprised of outlining a dozen of data analytic enablers to enrich the set of ontologies and its allied set of semantic engineering Tech: spark, R, python, owl, sparql. 02/08-08/13 Web/Software Developer Doctor tech. ltd. Pakistan • Design and develop multi-threaded applications and web development • Assessment of software requirements and specifications • Analysis, debugging, and code review • Writing multiple codes, and scripts to ensure cross-browser compliance Tech: Java, Dot Net, T-SQL, ADO.net, XML/JSON, .Net Framework and WebForms, PhP, MS SQL, T-SQL, ADO.net, ASP.NET, VB.6. 05/99 - 02/08Senior Engineer PAEC Pakistan • Assessment of software requirements and specifications • Network management and customized software development • Producing regular and ad hoc reports to agreed timescales as required • Mentoring team members in the development and technical progression Tech: PhP, MS SQL, T-SQL, ADO.net, ASP.NET, VB 6