Ketul Parikh Email & Phone Number

- Developed Cloud Infrastructure on AWS- Built CI/CD using Jenkins and GitHub actions- Used the following AWS service i. Deployed services on AWS EKS ii. Used AWS S3, SQS, SES, and EFS for our ETL pipeline iii. Used AWS CloudWatch (X-ray, dashboards, metrics, alarms) for the observation and monitoring iv. Used IAM and CloudTrail for Security and Anomaly Detection- Built ETL pipeline using NodeJs- Worked on MongoDB and Snowflake for the data processing

Worked as a backend developer.• Technology Used: Python, Django, Swagger, Postman• Developed REST API for OTT platform configuration• Used video encoding and decoding libraries in Python• Developed Unit test cases

Ahmedabad, Gujarat, India

• Directed a team specializing in advanced attribute and garment size detection.• Implemented a dynamic model refinement and deployment pipeline, accommodating weekly data updates.• Crafted a customized pipeline to generate choice-based attributes based on detected poses, colors, and garments,in line with client requirements using Python and MongoDB.• Took the lead in developing internal payment management software, leveraging Node.js.

India

• Developing CV-based attribute (Color) classification for further converting input images to a specific color.• Implemented CV-based attribute detection to eliminate shadows, adjust skin tones, and enhance background.• Collaborated with the QC team to create a correction dashboard for client attribute adjustments.• Successfully integrated with Google Cloud services for enhanced functionality and performance.

Title: Object recognition Using Deep Neural Network (DNN)Description: Given an image which has a single object, use Deep Neural Network approach to recognizing the object. Our goal is to learn the object model using the large dataset in an unsupervised way. We then obtain the features of labeled objects using this model. A supervised training approach is then used to solve the classification problem.To do so, I have done some prerequisites and introductory work:a) Supervised learning - Regression models and neural network on IRIS and WINE dataset.b) Unsupervised learning - Implemented Sparse Auto-encoders on MNIST and CIFAR dataset.c) Train an auto-encoder for estimating a Super-resolved image from the LR-HR dataset.d) Image Inpainting using Sparse Auto - encoder.e) For pre-training network implemented Restricted Boltzmann Machine (RBM).f) Implemented leNet-5 Convolutional Neural Network.g) Studied and implemented major UFLDL exercises.

Teaching Assistance:- Discrete Mathematics (Tutorial Session)- Digital Logic Design (Lab Assistance)- Introduction to Programming (Lab Assistance)

-To provide best sort of vehicle to handy‐cap people. Apart from physical movement they can move wheel chair by speaking few words. Made the sample model,operated on 12 volts DC power supply- Analysed the integrated potential of ICs and sensors controlled by PIC16F877.- Developed a deep understanding of the Vocal command technology and its conversion to electrical signals.- Undertook the application of Easy Voice Recognize Module to convert human voice to electrical signals.- Familiarized with the ISIS Professional 7 Proteus tool which served the base of the entire project idea.- Successfully developed VOICE CONTROLLED WHEELCHAIR with two modes of the application.- Was awarded the highest ‘AA’ grade by Gujarat Technological University for the project.

Master Of Technology (M.Tech.), Machine Intelligence

Bachelor Of Technology (B.Tech.), Electronics And Communications Engineering

Activities and Societies: Coordinator at XENESIS - National Level Tech-festival. Volunteer at XENESIS - National Level Tech-festival.

Intermediate/+2 (Science), Agriculture, Agriculture Operations, And Related Sciences

Schooling, Schooling

Activities and Societies: - Drawing competition - Dance participation -