HARATHI SURYA PATCHIPALA

Project: Lineage Datamart

Description: Lineage Datamart serves as the curated relational database for tracking data movements and transformations across the enterprise, ensuring a robust data ecosystem.

Roles and Responsibilities:

• Architected an AWS micro-services framework utilizing multiple Lambdas with SQS queues and DynamoDB datastores.
• Worked with product managers to get project requirements and spearheaded the overall project leadership
• Proficient in handling data within cloud databases such as AWS DynamoDB
• Skilled in querying relational databases including SQL and Snowflake
• Proficient in Scala for application development
• Implemented integration tests to detect defects during real-time pipeline execution
• Designed data structures to efficiently store curated data within the database
• Experienced in CI/CD pipelines for testing and deploying microservices to production

Project: Data Monitoring And Alerting

Description: Data Monitoring and Alerting system tracks dataset timeliness and quality, promptly alerting producers and consumers of any deviations from defined data quality rule.

Roles and Responsibilities:

• Contributed to the architectural design of the monitoring and alerting system.
• Developed Scala microservices to assess dataset quality and timeliness against predefined rules.
• Utilized JEXL (Java Expression Language) for runtime evaluation of expressions and scripts.
• Implemented integration tests to identify real-time pipeline issues.
• Managed multiple DynamoDB tables to store design-time and runtime data across microservices.

• Object Detection - Used Semantic segmentation model (FCN) to localize Handwritten text, signatures, hand checked areas in a scanned document
• Image Denoising - Used deep learning algorithms like CNNs to classify noise in an image and UNet to remove the noise from image
• Simulation of different noises in the image to create dataset for image denoising
• Image preprocessing using python Opencv and Numpy
• Handwriting OCR - Used CNNs, LSTMs to digitize handwritten text in a document
• Transfer learning to finetune trained models for customized data
• Experience using google tesseract to extract machine printed text in a document and finetuning of tesseract to work on specific type of documents
• Document text classification using NLP
• Entity extraction in a document using deep learning based NLP models
• Worked with stakeholders to develop roadmaps based on impact, effort and test coordinations.

• Developed python libraries for AWS S3 read / write operations using boto library
• Python code optimizations for better performance
• Debugging and fixing bugs in the given code base
• Experience with different file formats like JSON, CSV etc.