About Me

A Data Science Enthusiast and a once Senior Software Engineer, currently based out of München, I have an experience of close to 4 years as a Full-Stack Developer. During my early days on the job, I acquired an interest in data-driven Deep-Learning (and ML) based applications that drive most sectors of industry today. Taking some experiences and inspiration from a few solutions I was able to build as part of my first job, I have since pursued and completed my master's degree in the Data-Science Specialization at Universität Paderborn. As a former Machine Learning Intern and from my Master's Thesis in the area of Sentiment Analysis and NLP, I wish to further grow my career as a Data Scientist and hence, it is my desire to improve my instincts and thereby widen my understanding of the practices and methods that a data-driven solution would be built upon.

Contact Details

Pavan Kumar Sheshanarayana
Limesstraße 103
München 81243
(+49)176-34323676

Download CV

Education

Universität Paderborn

Masters in Computer Science• April 2019 - September 2022

  • Master Thesis (with BMW Group):
    "Cross-Domain Aspect-Based Sentiment Analysis with Multimodal Sources"
    • Deep learning models capable of extracting Aspect terms and sentiment developed as part of the core task.

    • Macro-F1 performances of the models increased in the range of 25-29% over the baseline with the application of semisupervised methods.

    • The models evaluated for their cross-domain adaptability scored Micro-F1 scores in the ranges of 80-95% across 4 of the overall 5 domains demonstrating the capability of domain adaptation in terms of sentiment

  • Other Modules Completed:
    • - Statistical and Natural language Processing.
    • - Computational Argumentation using Machine Learning and NLP.
    • - Online and Adaptive Machine Learning
    • - Interactive Data Visualization.
    • - Intelligence in Embedded systems.
  • Project Work:
    • Knowledge Graph and Entity Summarization using Linkbased Algorithms such as LinkSUM, SALSA and HITS

    Vidyavardhaka College of Engineering

    Bachelor of Engineering in Computer Science • September 2012 – August 2016

  • Thesis Project (with Webnock Technologies)
    • An Adserver based off of 'Revive' was developed by a team of us, for the organisation, Webnock Technologies. This was then used by the said organisation to manage the ad-campaigns, portfolios for their clients. The adserver was able to deliver numerous types of ads using a host of delivery methods to the intended websites and in-turn was able to capture the impressions, clicks and conversion statistics to help better manage the campaigns.

    Work

    Machine Learning Intern

    BMW Group • May 2021 - October 2021

    Involved in tasks ranging from data wrangling to Deep learning modeling and Visualization as part of metrics. All the tasks were put together to be a part of a web app capable of demonstrating the scope of building a multi-modalbased Sentiment Analysis.

    To analyze sentiment and recognize emotion as part of the above, two different models (NLP and Computer visionbased) from different modalities were trained and incorporated.

    Software Developer

    Myview systems • December 2019 - April 2021

    Worked as a full stack developer in integrating cloudconvert to the product 'xom' and was also involved in the maintenance of the xom API with simultaneous development of its client application.

    Senior Software Engineer

    Happiest Minds• October 2016 - January 2019

    Development of a few rich Key Performance Indicators (KPIs) for a university entity on a 'Student life cycle' from Cassandra Datalake.

    Development of a solution for a student entity to guide him on his four-year study plans (link-based Analytical algorithms employed, an ongoing POC at the time of departing with the company).

    Analytical showcase comprising various visualizations and metrics from multiple backend KPI microservices brought up as dashboard widgets.

    Have a look at my CV for complete details: Download CV

    Skills

    Starting my career in Full-Stack Development in an analytics team, I developed a keen interest in getting into developing Data-driven solutions. I have since completed the above-mentioned modules in my Master's Studies at Universität Paderborn. I have also completed the Deep Learning course (from Coursera and Andrew Ng) and recently finished my Thesis in the area of Aspect-based Sentiment Analysis & NLP.

    Along with building solutions with python over the last couple of years, I also bring my experiences in Java and Angular (& JavaScript) from my earlier years in the Industry. I am also familiar with the cloud services such as AWS and Azure. That said, I believe I can quickly familiarize myself with changing needs and norms of the Field since adaptability is of great essence. Hence, with close to two years of experience in all things NLP and further fundamental acquaintances in the other areas of Deep Learning (and ML), I believe and hope that I can build on the essential skill-sets that are integral to Data-driven solutions.

    Complete info regarding the tech stacks that I have worked on over the years can be found in my CV here!

    • Python (Pytorch, TensorFlow, scikit-learn, pandas & numpy)
    • Java (Spring)
    • Angular & Javascript

    Get In Touch.