About
Activity
-
Most agentic data pipelines fail in production because the LLM is doing too much. Not because the model is bad. Because it's in the wrong place. I…
Most agentic data pipelines fail in production because the LLM is doing too much. Not because the model is bad. Because it's in the wrong place. I…
Shared by Sushidhar Jayaraman
-
Data engineers spend most of their time building pipelines. But every now and then, a problem finds you that has nothing to do with pipelines. And…
Data engineers spend most of their time building pipelines. But every now and then, a problem finds you that has nothing to do with pipelines. And…
Liked by Sushidhar Jayaraman
-
Data engineers spend most of their time building pipelines. But every now and then, a problem finds you that has nothing to do with pipelines. And…
Data engineers spend most of their time building pipelines. But every now and then, a problem finds you that has nothing to do with pipelines. And…
Shared by Sushidhar Jayaraman
Experience & Education
Publications
-
Chaos based Shift Keying for Covert Communication
IEEE Xplore
-
CSK Based Secret Communication
International Journal of Engineering and Technology (IJET)
Courses
-
APPLIED BUSINESS AND ECONOMICS DATA ANALYSIS
-
-
Advanced Business Statistics
-
-
Analysis and Design
-
-
Applied Database Management Systems
-
-
Big Data Analytics
-
-
Computer Architecture
-
-
Data Mining
-
-
Data Science
-
-
Data Structures
-
-
Data Warehousing
-
-
Digital Signal Processing
-
-
Embedded Systems
-
-
Microprocessors and Microcontrollers
-
-
Mobile and Wireless Communication
-
-
Probability and Random Process
-
-
Signals and Systems
-
-
Social Network Analysis
-
-
Statistical Theory of Communication
-
-
VLSI Design
-
-
Web and Social Analysis
-
Projects
-
Pokémon Go App – Prediction of User Ratings for iOS and android versions
• Wrote R program to build regression models like Linear Regression, Ridge and Lasso to predict user ratings.
• Performed Deep Learning in app’s screenshot to identify Pokémon images using Tensorflow.
-
Kaggle Competition – Event Recommendation Engine
• Built an event recommendation model to recommend which event users will be interested in.
• Used Apache Hadoop to store 10 million Unstructured data and wrote PySpark Program to build item-based collaborative filtering
-
Kaggle Competition – IBM HR Analytics – Prediction of Employee Attrition Rate
• Cleaned, interpreted and analyzed data using seaborn, matplotlib to identify key metrics and created visually impactful dashboards in Tableau and Excel for logistics calculation and department attrition analysis
• Wrote Python program to build classification models like Naïve Bayes, Decision Tree, Support Vector Machine and Random Forest feature reduction to predict Attrition Rate.
-
Twitter Sentiment Analysis
• Analyzed people's sentiment and topics about the president Donald J. Trump's new administration.
• Collected 5 million tweets using Twython Streaming API
• Wrote Python program to analyze the sentiments and built Non-Negative Matrix Factorization (NMF) model from Scikit-Learn and Latent Dirichlet Allocation (LDA) model from GENSIM for Topic Selection.
-
Business Statistics – Dallas Housing Prices
-
• Analyzed what factors determines the price of houses in Dallas using the statistical methods like ANNOVA, Simple Linear Regression and Chi Square.
• Tools Used: SPSS, Excel.
-
Data Mining – Speed Dating Experiment
-
• Evaluated what attributes influence the selection of a romantic partner using data mining algorithms like J48, Naïve Bayes, PART, IBk, Decision Table and analyzed the results.
• Tools Used: Weka, Excel.
-
Developed a desktop application using JavaFX and J2EE
-
• Developed an e-commerce application for admin and customer role to perform transaction.
• Followed MVC architecture pattern and implemented a DAO layer to handle data from MySQL database.
• Designed a database and implemented CRUD functionalities to all entities. -
Chaos based Shift Keying for Covert Communication
-
A complete Information Security Management Scheme through obscurity was successfully built and tested by generating two chaotic waves using the famous Chua's circuit. Representing 1's and 0's as two asymmetric waves was the core idea of this project, through which the intended information which the recipient needs from the sender was successfully hidden. Statistical analysis was done to analyse the strength of this system. This project resulted in one IEEE conference paper and one journal…
A complete Information Security Management Scheme through obscurity was successfully built and tested by generating two chaotic waves using the famous Chua's circuit. Representing 1's and 0's as two asymmetric waves was the core idea of this project, through which the intended information which the recipient needs from the sender was successfully hidden. Statistical analysis was done to analyse the strength of this system. This project resulted in one IEEE conference paper and one journal publication.
This worked formed my bachelor's thesis and I played a significant role of managing all the resultant data and analysing them. I was also incharge of maintaining the project flow and point of contact with the project supervisor
Honors & Awards
-
Lead for Black Duck Analysis
-
Lead for Black Duck Analysis in Data transformer team and completed all the fixes needed for the product to be released in market.
-
Lead for Security COE
-
Lead for Security COE (Centre of Excellence team), identified and provided solutions and fixed all the bugs which lead for the product to be released in market with bug free.
-
On the spot award
-
Selected as a Winner out of 100 employees for “on the spot award” for obtaining “PLATFORM CERTIFICATION” for Data Transformer which is required for product to be released in market.
-
Received TCS Gems
-
Received TCS Gems appreciation from Module Lead for possessing good innovative skills and coming up with clear implementation, proactive and improving the team’s productivity
-
Star of Learners Group
-
‘Star of Learners Group’ – Highest award in TCS training program for designing a LOGO for Insurance based concept and got appreciation from Program Manager.
Test Scores
-
Masters GPA
Score: 3.82/4.0
-
B-Tech CGPA
Score: 7.2/10
Languages
-
English
Full professional proficiency
-
Tamil
Native or bilingual proficiency
More activity by Sushidhar
-
Today Roku surpassed 100 million streaming households around the world. When we started Roku, the goal was simple… make TV better for everyone…
Today Roku surpassed 100 million streaming households around the world. When we started Roku, the goal was simple… make TV better for everyone…
Liked by Sushidhar Jayaraman
-
After an incredible four-year journey with Roku, I have decided to turn the page and embark on a new chapter. During my time there, the team and I…
After an incredible four-year journey with Roku, I have decided to turn the page and embark on a new chapter. During my time there, the team and I…
Liked by Sushidhar Jayaraman
-
Excited to share my latest certification from Amazon Web Services (AWS) - AWS DevOps Professional
Excited to share my latest certification from Amazon Web Services (AWS) - AWS DevOps Professional
Liked by Sushidhar Jayaraman
-
I absolutely bombed an interview last Friday... Jumbled up my words, got lost in the middle of my answers, became so flustered over a slightly more…
I absolutely bombed an interview last Friday... Jumbled up my words, got lost in the middle of my answers, became so flustered over a slightly more…
Liked by Sushidhar Jayaraman
Other similar profiles
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content