Engineer Data

ARAVIND VARAGINA BANAKAR

**-***, ***** *, ********* Park, E-mail: *.*.*******@*****.***

Singapore - 118998 https://www.linkedin.com/in/aravind-v-b-80582380

Contact No: +65 93701008

OBJECTIVE:

I want work in the demanding and challenging environment where I can share my skills to create something valuable to organization and grow with it as it grows.

EDUCATION:

Master of Science in Electrical Engineering (Specialization Computer Engineering)

National University of Singapore, Singapore

Bachelor of Engineering in Electronics and Communication Engineering

The National Institute of Engineering, India CGPA: 8.7/10 AUG 2010 – May 2014

TECHNICAL SKILLS:

Tools, Libraries: RapidMiner, TensorFlow, Keras, SciKit, Tableau, JIRA, Eclipse, CM Synergy, GitHub.

Simulation Packages: Python packages like PANDA, NumPy, SciPy, TensorFlow, SciKit learn, OpenCV, MATLAB.

Languages: Python, R, C, SQL, Hadoop, Spark(basics), Embedded C, VHDL and Arduino.

Relevant Courses: Management of Industrial R&D, Pattern Recognition, Neural Network, Computer vision, Real-Time Systems, Advance Computer Networks, Multiprocessor systems, wireless and Sensor networks, Embedded Hardware Design.

WORK EXPERIENCE: R&D Graduate Software Engineer at Delphi Automotive Systems- Technical Centre India. -- 2 years of Experience July 2014 – July 2016

As software engineer I designed, developed, tested and delivered the software for Electronic Control Units and Powertrain Control Units using C, C++ and data structures. I was involved in multiple projects and worked with multiple client such as GM and Volvo. With Extensive research and analysis, I have developed Automation scripts for time critical RTOS tasks and functions using Python. I was having constant interaction with customers to understand their requirements and deliver the products that satisfy their needs. I was fortunate to be part of Patent review board which included higher management and technical gurus and contributed in launching in house idea platform. I had developed the model to suggest product enhancement and new feature additions based on the data collected from these users. I have filed a record of innovations which will be taken up for patenting.This got me appreciations from higher management and some rewards.

Internships:

Data Scientist Intern: @ TrySteve company.

TrySteve is a business and employment oriented platform like LinkedIn. In this position, I was responsible for providing the insights regarding the job seekers patterns and their trend in using the platform so that the marketing team and other core software development team can concentrate on improving the service and attract more job seeker and employers into the platform. Also I have worked on providing them a recommendation engines based on users behavior.

Trainee Engineer – Digital Switching Systems at RTTC, Mysore March – May 2013

RTTC – Regional Telecommunication and Training Centre focus on research area of modern digital switching systems and telecommunication systems. I implemented the system that could do the auto switching of regional wired telephony systems, account management and billing system.

Deep Learning Course by Google:

This online course was conducted by google scientific division for machine learning and data science in Udacity. It is 3 months self-paced learning course which gave insights about deep machine algorithms and how can it be used in data analytics /science projects. I have done several capstone projects using CNN, RNN and Deep neural networks using TensorFlow. This improves the model accuracy and helps in achieving better decision taking models rather than using simple machine learning algorithms.

Quantitative Risk Modeling (QRM):

This professional certification was thought by celebrated author and well known quantitative modeler Alex McNeil in DataCamp online certification platform. In this I have learnt some of the advanced quantitative modeling skills and done several projects on real world problems/data.

PROJECTS:

Stock Market Financial data prediction and Product Suggestion:

Developed a recurrent neural network(RNN)- Long Short-Term Memory(LSTM) model for predicting Gold price rise/ fall and another model to predict future closing price based on 10 years of prior data scrapped from different financial data sources such as yahoo, CME, quandl etc. Data was cleaned using different techniques using pandas. PCA is applied on features to select most prominent features using sklearn APIs. Skopt is used to get the hyper optimized parameter values to use in the model. Model can predict the trends fairly accurate and this is visualized using matplot libraries and Tableau. Based on the outcomes the financial products were suggested to customers.

Classification of Spam and Non-Spam Emails

Designing a Logistic Regression Classifier, Naïve Bayes (Beta-Bernoulli classifier and multinomial Gaussian classifier) and K nearest neighborhood classifier for email classification along with performance improvement by pre-processing of the data set.

Detection of breast cancer

Designed a MLP neural network that classifies the breast cancer cells. Models is developed using SVM to classify the cancerous cells and bagging techniques to cluster them. Very high accuracy has been achieved. Preprocessing of data and feature engineering has been done for faster and accurate model.

Credit card fraud detection:

Fraud detection can be done based on user spending behavior and sudden changes in trends. Data is collected from Kaggle sources and fraud detection model is developed using ensembled bagging and random forest algorithms SciKit Learn. This model is simple and effective. I could achieve good accuracy score (0.899). More advanced algorithmic techniques can be used to improve the scores and make it production level models.

Content Based recommendation engine

Developed simple production level content based product recommendation engine that recommends the products based on text descriptions. In this model, TF-IDF matrix will be generated based on the content and 100 related products will be selected using linear kernel much like cosine similarity and those products will be sorted using scores and stored in a list. This can be used to display those products as suggestions.

Smart house using IoT Connected devices

Designed and developed smart house device using sensor tag and Raspberry Pi controller along with 3 other team mates. Sensor tag is connected to Raspberry pi using Bluetooth. Data is collected in real time and prediction and decision-making algorithm is run on this [Time series] data to predict different conditions and make decisions based on that. This Hardware –Software co-design is developed using java script to collect sensor data in csv file and run LSTM model (python script) to predict the future data and svm model is applied to classify the ensembled data set. Future work is running the prediction and decision taking algorithms on the cloud using streaming data on Azure or equivalent platforms.

PUBLICATIONS AND PATENTS:

Filed an ROI (Record of Invention) on safety system for bikers at Delphi Automotive Systems, is taken up for patenting (in US) after it cleared in India level patent review board. Feb – 2016

Presented a paper on titled “Knocking” Tech-Café annual technological event at Delphi Automotive Systems- Technical Centre India March – 2016

Presented a Paper titled “E-Wallets and its technology trends “at National Level Paper presentation event at National Institute of Engineering. March – 2014

ACHIEVEMENTS:

Awarded prestigious “President Award” in Indian Scouts and Guides Indian unit on May – 2010 and received it from honorable President of India.

Awarded with “Employee of the month” award (SPOT Awards) and “Young Turk” awards in Delphi Automotive Systems for couple of months (June, August 2015) for my contribution in Patent review board and other initiatives.

Awarded with “Excellence Award” for Technical expertise and for conducting series of technical learning sessions at Delphi Automotive Systems.

Contact this candidate