The University of Texas at Austin

Software Engineer

New York, New York, United States

Google

- Founding Engineer of the AWS WorkDocs team that built and maintained the various components of the backend service to power mobile & desktop clients for enterprise document authoring & sharing

-  Overhauled the Asynchronous work processing fleet to trigger periodic notifications and accommodate scheduled transactions for internet scale usage patterns

-  Designed & implemented a Timeline View API to show actions performed by users pertaining to document authoring and collaboration within an organization

-  Designed & implemented Resource Sharing APIs to allow collaboration within an organization as well as external users with various levels of access controls and permissions

-  Gained broad knowledge of AWS data centers and applied that knowledge to solve real world physical constraints that have resulted from non-uniformly oversubscribed services offered by AWS

-  Overhauled the file synchronization protocol and client code to transfer documents more efficiently to quickly identify differences between data on users’ machines and the AWS WorkDocs Backend

-  Setup & maintain ETL processes to aggregate application usage data along with other data sourced from internal data warehouses at Amazon to allow product management team to gain better understanding customer behavior

-  Setup Redshift data warehouse for WorkDocs’ team to allow ad-hoc queries by leadership and other non-technical members of the staff to validate hypotheses around customer engagement

-  Improved infrastructure deployment & monitoring throughout the fleet creating alarms whenever a customer would have a bad experience due to increased latency or increased error rate

-  Within a few weeks of joining the team, received praise from management team: “Your communication rocks, definitely within top tier of software engineers” and “Your code review discussions raise the standards and have positively impacted team culture”

Amazon

Software Development Engineer - AWS

- Developed MapReduce jobs to calculate statistics about interesting events against web scale data streams. Designed the schema to persist these statistics in a NoSQL database for quick retrieval for any given time period, for any granularity, with heavy constraints on computation resources.

-  Authored a Maven plugin to reduce development time of new MapReduce jobs from weeks to days by simplifying job scheduling logic, abstracting details about cluster infrastructure and creating sandboxed environments with access to shared data sets.

-  Migrated existing ETL processes to use this plugin in production environments to reduce costs associated with maintainability.

-  Improve test coverage for Mappers, Reducers and Database Access Objects by writing unit tests against in‐memory versions of the real infrastructure to test the behavior of edge cases without requiring connection to shared resources.

-  Integrated the newly available data assets with legacy apps used by the sales teams to provide data regarding Bazaarvoice’s product usage that was not readily accessible before and is critical in lowering client attrition rates.

-  Coached engineering and non‐engineering teams to leverage the existing infrastructure to extract insights on demand from existing data assets.

Bazaarvoice

Software Engineer, Big Data Analytics

Ravel provides deep analytics tools that are capable of accurately discovering knowledge from databases, Hadoop, graph data, and sets of disconnected data. Ravel's core product, Golden Orb, provides graph-processing capabilities that enable companies to consume social network, machine network and other graph data sets on commodity hardware at significantly higher performance levels.

-  Developed distributed algorithms using Hadoop for calculating metrics on online communities.

-  Applied graph algorithms such as PageRank to discover influence of individual users within large social graphs.

-  Provisioned and bootstrapped servers within EC2 to create development and production Hadoop environments.

-  Automated the creation and the backing up of Hadoop and Cassandra clusters.

-  Developed interdependent scheduling and synchronization process for various long running work flows utilizing Azkaban.

-  Developed a distributed web app with load balancer to share derived data with clients through a web API.

-  Worked with the product team to offer quality assurance of Ravel's core product, GoldenOrb, a computation framework which simplifies bulk synchronous parallel graph processing, modeled after Google’s Pregel

-  Built web software using Google Web Toolkit and other open source technologies to let clients visualize and analyze time-series data.

-  Built code generation patterns to develop non-trivial crawling agents for fortune 1000 company websites.

Amaan Penang