716-***-**** email@example.com LinkedIn: Nagadeesh Nagaraja
MASTER OF SCIENCE DECEMBER-2017 UNIVERSITY AT BUFFALO, THE STATE UNIVERSITY OF NEW YORK.
Major: Computer Science GPA: 3.62/4.0
BACHELOR OF TECHNOLOGY MAY-2014 AMRITA UNIVERSITY
Major: Electronics and Communication GPA: 8.18/10.0 Skills & Abilities
Java, Python, MATLAB and C/C++.
TOOLS AND FRAMEWORKS:
CANalyser, CANape, SVN, MKS, Rational DOORS, Microsoft SQL Server, SSIS, Hadoop, Tableau, Pydev, Keras, Django, JSON, HTML ROS, Eclipse, RVIZ, VREP simulator, GIT, TEDESO, AWS/GCS- cloud computing platforms,. Experience
INTERN SIEMENS CORPORATE TECHNOLOGY, PRINCETON (NJ) JUNE 2017 – AUG 2017
Worked on development of a prototype to automate Model Based Testing of any Cyber Physical Systems (CPS).
Developed a simulated model (using VREP Simulator) and a controller (using Python), to control this model. Also built a framework to test the simulated model (using Java).
SOFTWARE ENGINEER ROBERT BOSCH ENGINEERING AND BUSINESS SOLUTION, INDIA JULY 2014 – JULY 2016
Worked as a Software Engineer for Application development of CAN communications. Worked on message transmission and reception for Adaptive cruise control system.
Was responsible for CAN communication monitoring, have worked on CAN communication-(PDU extraction) and Timeout, Checksum Alive-counter and Bus-off monitoring for CAN messages received and transmitted. This involved intensive working on CANalyser and CANape and Oscilloscope.
Worked on configuring BSW on AUTOSAR 3.1.4 version, for CAN communication stack using ARXML and CUBEC tool.
Briefly worked on CANSM Bus-off recovery State Machine, canIF, com and Pdur layers.
Handled the software development of 4 projects and was co-coordinator for testing the Application software for communication module and worked primarily with C language.
INTERN INDIAN SPACE RESEARCH ORGANIZATION, SATELLITE CENTER (ISAC), INDIA JANUARY 2014 – APRIL 2014
Project: Autonomous navigation of robot.
Developed an algorithm to combine the data from IR sensor, edge detection and the Gaussian mixture distribution (feature as pixels) to detect obstacles in the path of the robot. Achieved real-time obstacle detection, with 4 frames/sec detection rate.
Used ATmega 2560 series microcontroller to control the robot dynamics, DC motor, IR sensor and servo motors and used ZigBee communication to communicate with a PC.
Used MATLAB for Image Processing and Video was transmitted wirelessly through a dedicated transmitter.
Algorithm on object detection using image processing was published in conference: Nagadeesh N, Sanjeev Kaushik R et. Al., “Image Feature Based Navigation of Omni Directional Robot“, in: International Conference on Computer Science and Applications
(WCECS 2014), 22-24 October, 2014, San Francisco, USA, pp.516-520. Technical Projects
LOCALIZATION AND PATH FINDING ALGORITHMS ROBOTIC ALGORITHMS
Implemented ‘A*’ and ‘Grid Localization using Bayes Filter’ algorithm on the real data collected from the robot. Used ROS and C++ to implement the algorithms. For the localization task, the input was taken from actual movement of the robot, and achieved accurate results in terms of localization of robot position. TEXAS INDIA ANALOG DESIGN CONTEST.
Developed a prototype to use communication to control the head light of on-coming vehicles to lower the intensity of the head light
The communication was developed on ZIGBEE transceivers using MSP430 series microcontroller.
Our team was selected to participate in quarter-finals of the competition and later was selected to present a poster presentation at TI India Educators’ Conference-2013. Presented ‘poster paper’ titled “Auto Dim of head light”. 1
3 AXIS ROBOTIC ARM.
Developed a simple model and the controller for 3 axis robotic arm using servo motors and ATmega328 microcontroller.
The robotic arm was able to reach the destination provided with in maximum error of 0.5cm. Input to the robot was in Cartesian Co-ordinate, and entered on a PC console. The input was transmitted to the controller using UART communication, and the servo motor angles were calculated using simple trigonometry. GEODEMOGRAPHIC SEGMENTATION DATA SCIENCE.
Develop a predictive churn model for a customer dataset of a bank. Perform ETL process-Clean the raw data, upload it to database using SSIS and prepare the data through merging, transformation, elimination and dummy variable. Analyze the data through AB test and visualization (Tableau).
Develop predictive model through logistic regression and Back Elimination process (Adjusted R2). Test the data and compared the performance through CAP curve.
SONG LYRICS GENERATION USING LSTM DEEP/MACHINE LEARNING Generate song lyrics that are similar in style to that of a given artist, using the LSTM network. Trained our network on NVIDIA Tesla K80 GPU (using Google cloud services), network learned to form words, to capitalize, basic punctuation, and to generate verses of appropriate length. It produced varying lengths of songs- 500 to 4000 characters.
Implemented several classic algorithms to test and develop my understanding of the ML algorithms:
Comparison of neural networks on the classic hand written image and AT&T face database (subset) data set. Achieved accuracy of 95.5% on MNIST data set, and 82% on AT&T data set. Also worked on linear, logistic, ridge and lasso regressions. MEAN SHIFT BASED IMAGE SEGMENTATION COMPUTER VISION
Implemented a mean shift based algorithm to segment the color image into regions in the given pictures. Achieved the fastest segmentation time in the class of 110 people.
“Amrita-TIDE Innovation Awards”. Inter-college competition to bring out innovative ideas that help the society. Won 2 awards in a single year.