RPA Developer, 08/2018 HeidelbergCement AG – Heidelberg, Germany
Now Analysing business usecases to automate and provide technical solutions to improve the business by saving time.
Robotic Process Automation of the business usecases.
Direct interact with Users to analyse various business usecases based on SAP and other tools.
Analysing construction business data to provide technical solution.
Data Scientist Intern, 07/2017 AGT International – Darmstadt, Germany
Feb 2018 Sports Analytics, Internet of Things: Exposure to working with Elastic search for data visualization and analysis Improved data analytics skills by developing analytics component for Basketball game Implemented general workflow to run machine learning algorithm(used mainly for action recognition) Experience in sports analytics with wearable sensor data (action recognition) Familiar with understanding, analysing and visualizing raw data using python libraries (Pandas and Matplotlib), Also data augmentation Working exposure with time series data, which includes segmentation and sliding window approaches Expertise in feature engineering and feature selection Very good in model selection (sklearn) Skilled and comfortable working with various Machine learning and deep learning models (linear models, trees, NNs) using Sklearn, and other deep learning frameworks Keras, Pytorch and Tensorflow very good experience in training deep neural network well versed with different evaluation metrics.
Always worked independently under pressure.
R and D Engineer, 01/2014 to 01/2016
Nokia Networks, Bangalore.
Network Management System, Netact: Framework design and development in Shell and Perl scripting; Robot Framework: - Upgrade with cloning Framework Development - Datasync methods for upgrade with cloning Worked in Agile method, Scrum Master.
Systems Engineer, 01/2011 to 01/2014 Tata Consultancy Services – Bangalore
Linux Administration and Shell script developer.
2018 August - Autonomous Indoor Robot Navigation Using Deep Reinforcement Learning, Now DFKI.
Researching on Robot Navigation System for indoor industrial environment.
Using advanced learning techniques like Deep reinforcement Learning to train an agent to navigate a Robot autonomously from its position to the goal without any collisions.
The industrial robot is Robotino, developed from Festo Didactic, which offers all the basic sensor configurations to interact with the environment.
Working on several sensor data to observe the environment to create the map and to navigate.
Combining Deep Learning and Reinforcement Learning to solve this problem efficiently.
Python is used as a programming language since it offers very good framework for Deep Leaning and Reinforcement Leaning.
Robotic Operating System (ROS) has been used as a framework to operate the Robot, to provide inputs and read data from the Robot.
2018 May - Document Image Dewarping using Deep Learning, Conference ICPRAM-2019, July DFKI.
This research work is a pre-processing step for OCR to remove line curl and page curl from the input document using Deep Learning.
Supervised training of image-to-image translation network to remove the document distortions.
Acquired good working experience with Generative Adversarial Networks, which are trained using synthetic data.
The method and the results have been submitted to a conference since results are better than other existing methods.
This experience brought be tremendous knowledge in project management, planning and ability to work independently to solve any scientific problem.
2017 May - Prediction of Desk Occupancy, DFKI.
August Prediction of number of people inside a shared office space based on channel state information (Python) Data Extraction and segmentation, SNR, fft features extraction Classify the sequence on Baseline models(Sklearn) Classify the sequence (raw data) using LSTM and conv-LSTM (Keras) Experiment with different window size, hyper parameters optimization and results evalua- tion.
Masters: Computer Science (Intelligent Systems), 2016 Technical University of Kaiserslautern - Kaiserslautern Computer Science (Intelligent Systems)
Bachelors of Engineering: Computer Science, 2011 Dr Ambedkar Institute of Technology - Bangalore, 75 Computer Science
Well-educated Artificial Intelligent candidate with demonstrated ability in software development and to deliver valuable insights from a sea of information. Well-versed in machine learning, data mining, data analytics and deep learning. Able to process new data quickly and communicate it effectively to any individual Selected Machine Learning, Deep Learning, Artificial Intelligence, Embedded Intelligence,
Coursework Collaborative Intelligence, Multimedia Data Mining, Document and Content Analysis
In the area of Artificial Intelligence Design, build, deploy Robotics and AI applications to solve real-world problems To work in Big Data and Cloud Technologies To experience with different kinds of real-time data like image, text, speech, video etc
English Fluent Deutsch Beginner(A1)
ROS, Gazebo, RobotinoSIM, Reinforcement Learning, SLAM and Path Planning Algorithms Machine Classification, Regression and Clustering. Good working experience with varieties of
Learning: data. Deep Simple NN, CNN, LSTM and Image-to-image Networks.
Interests In the area of Artificial Intelligence Design, build, deploy Robotics and AI applications to solve real-world problems To work in Big Data and Cloud Technologies To experience with different kinds of real-time data like image, text, speech, video etc