Mahine learning Engineer - NLP, ML prediction models,DNN, CNN, Python
Resume:
Musarath J Rahamathullah
Machine Learning Engineer
SUMMARY
As a skilled Machine Learning Professional, I have successfully translated business requirements into models that were continuously improved and delivered tangible results. My core competencies lie in leading machine learning projects end-to-end with goal of improving customer experience and driving business impact.
EDUCATION
Masters in Analytics
Major: Machine Learning
Bachelor in Technology
Major: Electronics & Communication
SKILLS
4years of experience in Machine Learning, NLP, NLU, Deep Learning, building ML pipelines and data engineering
AI Platforms: DataRobot, Snorkel AI, IBM Watson
Frameworks: Hadoop, PySpark,
Apache Hive, Apache Impala, HDFS
Programming Languages: Python, R, SQL
Analytics: SparkML, Numpy, Pandas, Scikit-Learn, NLTK, Spacy, keras, plotly, matplotlib
App/Web Frameworks: Flask, Streamlit, ShinyApp
ETL Tool: Talend
WEBSITE
RPubs: https://rpubs.com/musarath
GitHub: https://github.com/Musarath-Rahamathullah
LinkedIn: https://www.linkedin.com/in/musarath-jahan-r-b979944a/
CONTACT
PHONE:
EMAIL:
***********@*****.***
PROFESSIONAL SUMMARY
Having 8 years’ experience in Software industry, including 4years of experience as ML engineer with proven success in building successful NLP recommendation algorithms and predictive AI (Artificial Intelligence) models especially for Financial, Banking and Healthcare industries
Worked for Investor Bankers (in BOFA) to predict potential sell side private clients using advanced ML models
Worked with Constance to provide Sentiment Analysis on voice call recording using speech recognition APIs and NLU.
Responsible for Data Analysis and Data mining for leading courier services provider which helped building business recommendation.
Analog (POTC) and Digital (BRIC) interfacing board design and programming for Fiber in Local Loop system.
ACHIEVEMENTS:
Publication: My paper “Diagnosing skin cancer on early stages using deep learning approach, has been accepted for a poster presentation at the conference on “AI and Big Data in Cancer: From Innovation to Impact” for one of the reputed publication Elsevier Boston. https://www.linkedin.com/posts/harrisburg-university_huproud-activity-6627662822895562752-22E_
Publication: My paper “A Proactive Approach to Combating the Opioid Crisis using Machine Learning Techniques” was nominated for 6th International Conference on Health Informatics & Medical Systems and for publication in Springer Nature - Research Book Series. A Proactive Approach to Combating the Opioid Crisis Using Machine Learning Techniques SpringerLink
Codeathon: Finalist of AI Code Week in BofA for “Who Knows Who” project, based on RFM Model to build Client Network graph.
EXPERIENCE
Bank of America (Machine Learning Engineer, 6/20 – Current)
Specialist in Natural Language processing (NLP) and Natural Language Understanding (NLU) of unstructured text and ML models.
Responsible for designing, building and deploying the Sales Commentary Recommendation Model to recommend sales commentaries to each external client or contact in Bank of America.
Built Global Corporate Banking (GCB) Sell Side prediction model in DataRobot, to predict the likelihood of a Middle Market sized client to do an M&A Sale transaction in the next 12.
Build individual models using Snorkel to identify 15 various entities from the case study documents of GCIB to build a search engine.
Design lean proofs of concepts (POC) to answer targeted business questions. Explore and work with a wide range of proprietary, interesting AI platforms. Apply existing methods or develop new methods.
Harrisburg University of Science and Technology (Research Assistant, 5/19 – 6/20)
Performed advanced Convolution Neural Network models along with VGG16 on the dermatoscopic images of skin leisons, to detect skin cancer type.
Sentiment Analysis of voice call recording using speech recognition APIs and NLU.
Harrisburg University of Science and Technology (Data Scientist, 1/18 – 12/19)
Ensured the data quality, consistency, and integrity using Python Pandas.
Performed data preprocessing on messy data including imputation, normalization and scaling using scikit-learn.
Performed Exploratory Data Analysis (EDA) to visualize through various plots and graphs using matplotlib and seaborn library of python, and to understand and discover the patterns on the Data, understanding correlation in the features using heatmap, performed hypothesis testing to check significance of the features
Carried out feature selection by techniques such as feature importance, feature correlation and feature ranking with recursive feature elimination using scikit_learn.
Applied unsupervised techniques such as Hierarchical clustering and K-means for customer segmentation.
Developed a Churn model for the Human Resource team to increase the retention rate of employees
Built multi classification models based on Logistic Regression, Support Vector Machine (SVM), Random Forest, Ada boost and Gradient boosting using Python, Scikit-Learn to detect single Thorax disease from chest x-ray images with average f1-score of 0.92
Evaluated performance of classification models using F1 score and AUC of ROC and PR curve.
Performed Exploratory analysis on the monthly expenditure of a user and build dashboard to visualize and understand the trends and seasonality in the spending of the customer profiles using R and ShinyApps, decreased the travel expenses by 9% .
Performed NLP by using techniques like Word2Vec, FastText, Bag of Words, TF-IDF, LDA and ML algorithms for Twitter Analytics on businesses to extract insights about business on social media.
Developed an application for Topic Modelling of News text using LDA, Gensim and BERT, and analyzed the stocks based on the news.
Image Classification and Image processing using Tensorflow, Keras for Multilabel Skin Cancer detection, with Sensitivity of 96%.
Capgemini Consulting Limited (Senior Consultant, 8/09 - 2/12)
Served as a key member of software development team for the DTDC Global and DELL E-Commerce clients.
Analyzing user data, extracting business insights, building business recommendations, which increased the company’s revenue by 17%.
A/B statistical testing new product ideas and communicating findings to various stakeholders.
Identify and document data integration anomalies, then work with development to resolve issues.
Pinpointed previously undiscovered flaws in new DTC system prototype, resulting in formal commendation letter from the DTDC BOD and project star performance award from the company.
System validation and reporting of online space for dell.com, for purchasing dell products across the globe. And perform root cause analysis on results.
Siemens Information Systems Ltd (Associate Consultant, 5/08 - 7/09)
Performed system, and data interface validation for Client-Server based architecture, Power generation plant management system and development of proprietary tools.
Development of automated BI reporting tool using VBA for generating summary and reports of all projects, project status and financial details across all verticals in the organization
Created a comprehensive database for documentation of all phases of project and procedures to enable accurate replication and ensure compliance with standards. Accelerated average validation turn-around time by 25%.
Cognizant Technology Solutions (Programmer Analyst, 2/08 – 5/08)
Information gathering, generating various reports and validations for consumption of the Software Assurance (SA) and Volume Licensing (VL) e-learning benefits offered to Microsoft employees
Industrial Consultancy and Sponsored Research, IIT Chennai (Project Associate, 6/00 – 3/03)
Development of FILL (Fiber in Local Loop) system, to give Analog and Digital telephone interface communication to user, which is deployed in BSNL.
Hardware board designing, ADSP assembly language programming for physical, data and Network layers of OSI (Open Systems Interconnection) model, debugging and unit testing.
Analog (POTC) and Digital (BRIC) interfacing board design and programming.
VHDL programming
AI PROJECTS
Text Analysis using NLP and ML Twitter Analytics on businesses
Aspect-based sentiment analysis over a period of time to understand what people are talking about a company or brand on Twitter media. Designed a text classifier for topic labelling with 91% accuracy, to help identify topics every time someone shares something about a brand. Moreover, combined these topic classifiers with sentiment analysis models to develop a real-time tool for a company, to leverage actionable insights to drive businesses. Click to preview the report of PPP loan.
Data Visualization Exploratory Analysis of monthly expenses
Conducted exploratory analysis, by creating the interactive dashboards using ggplot, dygraphs, shinyapps packages in R, to get insight of expenses, using data collected over a 3-month period of floating expenses. As a result of the analysis, I was able to comprehend the top categories of my expenses. And the travel expenses dashboard, aided to keep track of expenses and save 12.5%. Click to view the storyboard.
Contact this candidate