DIVANSHU VERMA

Professional Projects:

• Design and developed RaaS on SharePoint.
• Contributed in Sensitivity labels https://learn.microsoft.com/en-us/purview/sensitivity-labels

Roles and responsibilities

• Project High level and low level design, estimation and timeline.
• Development (Coding, testing and operational support)
• Guiding SDE for subcomponent of projects.
• Mentor SDE in team.

Professional Projects.

• Enabled Health Cards payments for Amazon pharmacy insurance flow for US and Canada.
• Design and enable PHI compliance for Amazon pharmacy.

Roles and responsibilities

• Project High level and low level design, estimation and timeline.
• Development (Coding, testing and operational support)
• Guiding SDE1 for subcomponent of projects.
• Mentor SDE in team.

This team is responsible for processing all amazon retail and digital business concessions. I worked in multiple teams Example : Concessions Execution, Concessions Policies and Digital Devices , etc.

In the span of 2 years in this team, I got a chance worked on multiple projects like Online Return Service deprecation, Exception engine, Concessions service as a platform, Concession service deprecation, Policy migration which helped me understand, how to deal with business complex project also improve my bias for action and project driving skills. This challenged pushed me to learn new business quickly and also increased my understanding of the concessions workspace and its problems. Team had picked up the task of solving some complex problems like Exception engine, concession service modernizations, It's helped me to explore, learn new architecture like platform based architecture.

Exception engine

Project Summary: System which will suggest the concessions exception as per the customer concessions history and other factors.

Time Duration : 6 month

Accomplishments: We successfully launched the project in different marketplace and increased the concession saving by 13.5M in 6 months.

Concessions service modernizations

Project Summary: Concessions services are based on monolithic architecture. Which increase the move to production time. We took a goal of Concessions service modernizations.

Time Duration : 2 year

Accomplishments: We re-architecture the concessions service as micro service architecture. We build a concessions platform in which any business can build plugin's to enable their business.

Concessions policies optimization.

Project Summary: Concessions is using multiple policies as per the order type and issues. Engine which host these policies were legacy. We need to build the new policy engine and optimize the polices for faster development and execution time.

Time Duration : 6 months

Accomplishments: We build a new policy engine which is customer friendly and easy to change the policy. This build multiple tools to complete this migration in given time duration.

Online return service deprecation

Project Summary: We need to migrate from online return service to newer platform based architecture. There are multiple concessions workflows which are using ORS API. We need to understand the business logic (Which is complex because of wide verity of workflows and use case.) This system is central system of all amazon business concessions issuance). Example of concessions like refund, replacement and free replacement.

Time Duration: 6 month

Accomplishments: We are able to completely deprecate the service within given time by taking multiple biased decision. We deep dived into the code, found blast radios for every biased action and provided data point to our higher management to take the decision.

Kindle Japanese Launch (Amazon)

Project Summary: First time we launched Japanese book with in the all kindle market with new enhances format and features.
Word Segmentation enabled Japanese books in Kindle publishing pipeline. Support vertical reading order in books with ruby and other Japanese specific tags.
Time Duration: 6 Months
Accomplishments: More than 1M Japanese books are released in the Japanese marketplace. Also, enabled Japanese books conversion in live pipeline.(100 Books/day). Because of this project we are able to launch new Kindle devices in Japanese market place.

Kindle page number generation service.

Project Summary: This service is responsible for generating page numbers for kindle books.
Generate page numbers through publisher provided data, if publisher is not providing data we will generator page numbers through bunch of other mechanisms. Multiple Listeners for triggering service. Validation frame work for checking generated page numbers accuracy. (It should be more than 95% accurate)
Time Duration: 9 months
Accomplishments: After launch of this service we are able to provide page numbers in 83% of books with more than 95% accuracy.

Manual workflow for quality issues.

Project Summary: This workflow will be scheduled for quality issues reported by the automation, customers or any other source.

• Run automation workflow for book quality validation.
• Created a manual operator workflow for operators to work.
• Carry forward issues from previous revision to new revision of book.
• Reconciliation of the same issues which are reported by multiple resources.
• Report issues to the publisher.

Time Duration: 6 months
Accomplishments: We are able to resolve millions of quality issues which are creating customer impact during reading of the book as well as we are able to increase the sale of the books.

Quick Heal Antivirus Update Optimization

Project Summary: Existing system was sending 4 to 5 update/day with size of 200 MB/ update. I need to build a fully automated system which will be capable of generating 20 updates per day and the size of update should be minimum.

• Implemented Tier-1 Service which will generate the differential update.
• Test the updates for all virus data sets for validation.
• Client-side application to fetch and apply update.

Time Duration: 9 months
Accomplishments:

• Update size decrease by 90 percent.
• Sales increased by 30 percent.
• Customer satisfaction index increased by 30 percent.
• Server maintenance cost decrease by 40 percent.

Update Manager Data Optimizations.

Project Summary: Update manager is the application which manages an update process inside the enterprise edition. It downloads update from quick heal server once and serve all his client through that update. This manager works as an internal server inside enterprise. We need to redesign the update manager to manage minimum amount of data to update all clients.
Redesigned and implemented new Update manager to store minimum data to serve all clients.
Time Duration: 3 months
Accomplishments:

• Sales increased by 15 percent.
• Update time decreased by 10 percent.
• Customer satisfaction index increased by 10 percent.

Continuous integration and Automation.

I designed and implemented a system to reduce the overall development time, increase the visibility and tractability of development work. Automation of C code compilation on 9 different machines and automatic testing of generated binaries on multiple platforms.

After that package will be hand over to QA team for manual testing and to thread research and response team for malware verification. When they promote the build, it will be automatically sent to release team where one click update automation system generates the update. After manual approval of release team, update will be uploaded on servers. Released binary symbols will be uploaded on symbol server and binaries will be uploaded on binary server. Also integrated SVN repository for tractability of code commits and Redmine for traceability of issues allocated to developers.

Scanner optimizations:

Project Summary: Scanner is the module which scans files of system.
I worked on scanner performance improvement.
1) Scanner caching support: Local caching support to the scanner for performance optimization.
2) Scanner whitelisting support: Files, applications and installers whitelisting support to the scanner for performance optimization.