No experience
Employment Type:
Full time
Job Category:
Software Engineer (Infra - Data Platform)
(This job is no longer available)
TellApart | San Francisco, CA
Grad Date

Not sure what types of jobs you are interested in?

Explore Jobs
Based on Your Education

Follow This Company

Job Description

Software Engineer (Infra - Data Platform)


We own all of the technical underpinnings of the infrastructure team. Every production system we've ever built feeds into the data platform (where we have PB's of batch-processed data under management), and we are also responsible for the streaming infrastructure for TellApart as well. Our team is also very active in the open source community =)


2+ years of industry experience building large-scale, distributed, high-volume systems (with Python, Java, and/or C++) (internships count toward this requirement)
Good understanding of multi-threading, scaling, concurrency, caching, queuing, RPC, and distributed data storage
Ability to mentor others and thrive in a dynamic, fast-paced, collaborative, and high-growth start-up environment
Strong attention to detail with a predilection toward automating almost everything

Bonus Points:

Experience working with the Lambda Architecture and systems like Hadoop, MapReduce, HBase, Hive, Voldemort, Azkaban, Kafka, Spark, etc.
Architecting and deploying asynchronous work queues, high-volume storage systems, and high-throughput systems excites you
Active contributor / committer to a well-known open source project and/or interest to do so in the future
Demonstrated ability to excel in whatever you pursue (whether it's work, school, competitions, open source contributions, personal projects, etc.--you've always stood out and succeeded)

Current Focus:

Build, extend and maintain the key infrastructure that powers the TellApart platform
Build the next generation of high performance, highly available, distributed systems
Make significant architectural decisions about the next generation of TellApart infrastructure
Take ownership of key components of the architecture that power the TellApart suite of products

Recently Completed Projects:

Built an identity management platform which associates a single user identity across devices and supports an anonymous profile service to derive attributes like gender and purchase preferences from their user history

Built a data pipeline which ingests > 100M user activities per day and creates a cumulative user history for use in training models and building datasets for real-time usage

Built a pluggable monitoring platform that allow users and services to retrieve, compare and alert on metrics recorded in OpenTSDB, Cloudwatch and other metric stores

Built query system from Spark, Tachyon and Parquet that caches Terabytes of data in memory and increased our query speeds by 100x. We offered contributions back to all projects in the course of development

Built a data extraction framework to enable simple and consistent data extraction from complex structures across multiple Map/Reduce platforms such as Hadoop, Hive and Spark

We are now working out of Twitter's San Francisco office.

TellApart is an Equal Employment Opportunity and Affirmative Action Employer with a commitment to workplace diversity. All qualified individuals are welcome to apply. Employment with TellApart is based solely upon one's individual merit and qualifications directly related to professional competence. We do not discriminate on the basis of race, religion, color, sex, age, national origin, citizenship, or disability. And we will make all reasonable accommodations to meet our obligations under the Americans with Disabilities Act (ADA) and state disability laws.

About TellApart

TellApart is a technology company that helps online retailers advertise to potential and existing customers. Our mission is to enable e-commerce companies to develop deeper relationships with their best customers through more relevant marketing communications. One of TellApart's most widely used marketing services is a display ads application for retailers called remarketing.