vimal dharmalingam

Experienced professional in Data Science, Machine Learning Developement to improve the Business value.

singapore, Singapore

Summary

• Totally 4 years of work experience as Data architect, Data Scientist/Analytics in Advanced Analytics, Predictive Modelling - Banking, Industrial and Government sector and e-commerce and IOT.
• Rich work Experience in Advanced Analytics , Data Science with Machine Learning, Deep learning, Text Mining and NLP.
• Experience in integration and adoption of Advanced analytics insights into existing business processes by demonstrating the ability to communicate the business benefits provided by analytics insights.
• Deliver Data Lake, setup infrastructure to support ingestion of the data from multiple sources.
• Experience on handling structured, semi-Structured and un-structured data.and ETL processes
• Hands- on experience in building predictive models using Supervised and Unsupervised Machine learning algorithms: Regression, Classification, Clustering and Natural language processing (Text Mining/Sentiment Analysis).

Languages:

English, Hindi, Tamil

Favorite Python Packages:

Scikit learn, matplotlib, Keras, Numpy, pandas, Scipy, NLTK.

Experience

Work Experience

Data Scientist (Lead)

Government technological Agency

Singapore

Aug-01 to Present

• Implementing the AI solution to reach the business goals.
• Experience in Data flow from scratch to frondend deployment.
• Data table schema experience in Nosql and sql databases
• Experience in creating Data lake in AWS 
• Responsible and Owned for End to End Machine learning use cases and Solutions using Sagemaker.
• Extracting the Data from source and manipulation of raw context , Finding the undefined patterns in the data to reach the business goals.
• Automation/Deployment of Machine learning models using aws services
• Realtime Inference prediction using Lambda function and API Gateway. 
• Implemented training and production ready model using AWS machine learning services. 
• Built and automated the Analytics solutions to the business
• Incorporating ML/NLP logics to improving the performance of elastic (ELK stack) Search Engine
• Automated the Text similarity model to find the application of sensors using BERT similarity model with semantic meaning
• Implemented the Data mining and Data mapping processes using NLP techs.
• Closely worked with the Data Engineer/backend and recommended which field to be in what data format. Also helped him to extract additional fields necessary for implementing supervised machine learning model.
• Was SPOC in the team and coordinated many modules/tasks and was helping delivery manager in achieving the same.

----------------------------------------------------------------------------------------------------------

 

Data Scientist

 Hong Leong Bank Berhad,

 Malaysia

Jun-2018 to July 2019

• Responsible for End to End Modelling process and implement large scale solutions using Python, jupyter, Teradata SQL & Base SAS.
• Experience in Model development and deployment in Hadoop environment ,spark.
• Predict the Customer service calls using ARIMA & SARIMA models
• Built a model to identify Customers without PL has high propensity to buy PL (CASA & CC Base).
• Built a model to recommend merchants by understanding customers implicit & explicit interests by analyzing their historical behavior.
• Sentiment Analysis-Build a model to evaluate customer’s feedback from multiple sources (Internet Banking, Social Media (Facebook, Twitter, Chat bot) & Call Centre.
• Model is back tested by real time manual campaigns and compare the results of both (Response Rate of Manual Leads vs. Model Leads).
• Experience in Personal Loan, Mortgage Loan, CC, CASA, INSURANCE etc.
• Hands on experience on Attributes Selection for ML & Evaluating the Models.
• Experience on Chatbot (IBM Watson) and Text Analysis

Project1:

• Problem Statement: To find high propensity customers who will buy Personal Loan in near feature for Marketing Campaigns
• ML-Tech: Ensemble Models

      Solution: ensemble model have been giving the 4% better lift than manual campaigns.

Project2:

• Problem Statement: Recommend the most appropriate merchant based on customer behavior & spending pattern.
• ML-Model: Text Mining, TF-IDF Vectorizer, N-grams, Cosine similarity, Item based Collaborative Filtering

Solution: Using Item based collaborative Filtering Recommended the Merchants based on customer spending behavior

Project3:

• Problem Statement: Developed python Machine Learning model to evaluated Bank’s Chatbot (IBM Watson) responses
• ML-Tech: TF-IDF, cosine similarity

Solution: Helps bank to retrain the bot on products where it was not answering accurately.

 

Data Modeller/Analyst

Consolidated Construction Consortium Ltd,

Chennai/Hyd (Reliance)

Nov-2016 to Apr-2018

• Undergone intense class room training for Model building , data cleaning in construction department  and designing of evaluation metrics according to use case- CM,Gain lift ,ROC curve.

• Started as a Data Science  Freelancer and become a full time  Data Scientist.
• Using the historical sales data and demographic using customer profiling developed Prediction model on house sales for campaign events and developed customer portfolio for targeting during events.
• To up sell the apartments, identified the wealthy customers using propensity modelling also targeted customer with closure to payoff their existing loans
• Mining the data for statistical Modelling using excel, Preparing daily safety analysis report, Finding possible ways to improve effective input features.
• Understand the Data and pattern needs to be Identified then plan the architecture.Models are built to predict  accident zone functionalities.
• Worked on feature selection using various technique to reduce the dimensionality of the data. Reducing the underfitting and overfitting problems.
• Cleaning huge data using pandas according to the requirement ,converting the data to the suitable format  for the model  - Sampling ,Encoding, Principle Component Analysis.
• Prepared analysis report about Linear and Non linear algorthim performance for various construction usecase and excuted with realtime data for trail purpose. Deep learning output is compared with traditional algorthim reports

Tennaco Automative India Pvt Ltd

Hosur, Tamil Nadu

May-2015 to Oct 2015—6 Months

• Implementing the businees process flow and Data flow from scratch.
• Analysing the production  data and deliver the Actionable report to the Higher management.
• Forecasting the  future requirement and production rate based on the insights extracgted from the existing business data..
• Mining the data from source using excel, Preparing daily  analysis report, Finding possible ways to improve effective input features.

   

Skills

Amazon Web Services (AWS), Artificial Intelligence, Big Data, Data Science, Git, Keras, Machine Learning, SQL, SciPy, Spark