Akshi Bhardwaj

AWS Marketplace's Product feed design

• Ideated and designed the new product data S3 feed to sellers that got rid of hacky implementations in current system.
• Successfully leveraged Flyway, S3 & Redshift spectrum to get rid of Amazon's black box ETL tool - Hammerstone.
• New feed was not only operationally more manageable but enabled faster delivery (from daily to hourly) of product data feed to our sellers as a major tech dependency was eliminated.

Integrating Lambda UDFs with Ingestion system to enhance upstream data

• Researched how to maximize throughput while integrating lambda UDF (User defined functions) with current Redshift based ingestion system.
• Designed a generic proposal solution with parameters like timeouts, retries, concurrency settings that doesn't throttle upstream services.
• Delivery improved substantially along with standard hourly delivery of reports rather than waiting 1 day for the old jobs that run on accumulated data throughout the day.

French Tax reports' delivery automation

• Created a CD pipeline to ingest French tax data using S3 and Redshift, processed through Lambda and finally delivered to secure buckets using KMS.
• Yearly reports to French government tax authorities started dispatching through automated jobs.

Statistics Platform Conversion & Fragmentation design

• Designed a service which could handle ingesting ~2 GB files, fragmentize, process and even convert file types if necessary.
• Implemented the solution using Elastic container service instead of previous Lambda which only supported 2048 memory read.
• Also built failure handling system to avoid partial data writes, data duplicity to storage using buffer storages.

Altea Reservation Desktop

• Developed RESTful web-services using JAX-RS framework and MyBatis framework for SQL.
• Developed an Android mobile hybrid application (built using CORDOVA) with hardware (Beacons) plugins integration that consumes data from services and provides local app notifications.
• Migrated database from SQL relational database to MongoDB.
• Improved real time tracking and updates to user using Kafka jobs.