Yen P. Email & Phone Number

San Francisco, Ca, Us

Haus combines state-of-the-art causal inference and econometrics with experimentation to help brands make informed investment decisions in an increasingly privacy forward environment. We have significant traction with well-known and fast-growing brands; and are backed by Insight Partners and 01A Advisors, joined by prominent early stage investors that include Baseline Ventures, Haystack, Upside, Box Group, Fathom, and the Chainsmokers (via Mantis VC)!Haus made LinkedIn's Top 50 Startups for 2024:https://www.linkedin.com/pulse/linkedin-top-startups-2024-50-us-companies-rise-linkedin-news-hxote/Notes from Insight Partners, who led Haus' Series A:https://www.insightpartners.com/ideas/behind-the-investment-haus/Get in touch if you want to learn more: https://www.haus.io/careers

Lafayette, California, Us

San Francisco, Ca, Us

I support Data Platform and ML Engineering at Discord, which includes data infrastructure/tooling, experimentation platform; and ML infrastructure and engineering. We manage a data warehouse of 1000s of tables and 20+ petabytes of data (and growing). Responsible for data and ML infrastructure budget of ~$25M USD.- Built up the overall Data Platform + MLE org. from 7 in 2021 to 50+ persons, organized as 6 teams covering data infrastructure and services, experimentation platform, ML platform, and applied ML.- On the data/ML foundations side, we shipped and/or supported: data catalog with lineage with DataHub; integrated Fivetran for 3rd party data; an evolved ELT framework automating data warehouse table build-outs; event-stream "taps" to power event-driven user features and workflows; improved experimentation UX and metrics pipelines; data warehouse exports to live datastore (BigTable) for online usage of transformed data; user and community similarity embeddings to power ML-driven studies and products.- On the ML side, we shipped and/or supported: ML-powered products to fight spam and prevent harmful or illegal activities; ML-powered products to highlight engaging content via community feeds and notifications.- Successfully integrated team members from Sentropy acquisition in Summer 2021 -- at the time this was about a 30%+ team size increase in one shot.- Established cross-functional Data Governance Council comprising of leaders from Data, Security, and Legal departments.- Updated recruiting, interviewing, and onboarding practices.- Educate and establish ML engagements in key product development areas at Discord.- Scaled processes to build up and reinforce habits around operational excellence.- Related blog posts: https://discord.com/blog/building-delightful-notifications-using-mlhttps://discord.com/blog/how-discord-creates-insights-from-trillions-of-data-pointsPlatform: Python, Java, Scala, Rust, BigQuery, GCP, Scylla/Cassandra, Kubernetes.

San Francisco, California, Us

Sonder went public in January 2022 (NASDAQ: SOND) and was backed by a collection of top venture and equity management firms, including Fidelity, WestCap, Spark Capital, Greylock Partners, Greenoaks Capital.I led and supported engineering teams at Sonder responsible for distribution, pricing, payments, financial reporting, experience quality, and operational efficiency.- Transitioned engineering team to a more comprehensive remote orientation as a response to Covid and for the future: ways of working, process, and recruiting changes.- Work cross-functionally to clarify domains, missions, and relevant metrics, aligning the engineering organization as necessary.- Established Frontend and Backend Guilds to drive technical initiatives that cut across team domains.- Established twice-a-year Hackathons to foster innovation.- Define and provide ongoing guidance to Sonder’s technical vision and roadmap.Platform: Go, Python, Ruby, AWS, Postgres, SnowflakeDB, Kubernetes.

San Francisco, California, Us

DoorDash went public in December 2020 (NYSE: DASH) -- DoorDash is Y-Combinator alumni (2013) and was backed by top investment firms Coatue, Dragoneer, GIC, Kleiner Perkins, SoftBank, Temasek, and Sequoia.Led teams at DoorDash responsible for the core consumer experience and delivery reliability. During my time at DoorDash, we scaled from 60k to nearly 1 million deliveries per day. Team: scaled Consumer team from <20 to near 50 persons in 2018; refined execution process by organizing sprint planning, quarterly resource allocation, prioritization; worked with recruiting team to refine interview process: module development, rubrics, interviewer training. Product: launched the DashPass subscription program, leading to sustained improvements in volume and order frequency; supported point-of-sale integrations with marketplace, improving customer selection by adding thousands of national stores; integrated DoorDash functionality with Google Food Ordering, augmenting SEO/SEM with an organic channel for new customer acquisition; improved delivery quality and reduced support contacts by 30% in 2017. Technology: performance/reliability improvements to support 300% year-over-year volume growth; migration of search/ranking functionality from Python monolith to service-oriented ecosystem for greatly increased scale, reliability, and ease of experimentation; worked with team to define vision for future state of marketplace platform. Platform: AWS (Kinesis, SQS, Aurora, DynamoDB, ElasticSearch), Python, Java, Kotlin, Kubernetes.

Greenwich, Ct, Us

Consulted on client projects and incubated product ideas (https://www.producthunt.com/posts/okk).

Emeryville, Ca, Us

TubeMogul went public in 2014 (Nasdaq: TUBE) and was acquired by Adobe in late 2016.Led core development teams in the design, development, and support of the systems that drive TubeMogul's programmatic branding platform: real-time bidding (RTB) ad systems, real-time analytics pipelines, and user targeting systems. Heavily involved in both technical strategy and the scaling up of the engineering team during a critical phase of growth. We built systems that experienced extremely high growth (5k to 200k requests per second in 18 months), supported yearly doublings of revenue to hundreds of millions, and which now do millions of requests per second powering Adobe's Marketing Cloud.Platform: Java, Kafka, Storm/Trident, Couchbase, Vertica, Aerospike, ZooKeeper; custom frameworks for data processing and distributed systems management (providing what services like Consul and parts of the Kubernetes ecosystem provide today).

Emeryville, Ca, Us

Lead role in the design and implementation of TubeMogul's video syndication product: a message-based system designed to dynamically scale on a cloud infrastructure. Designed and built workflow framework to coordinate asynchronous tasks across nodes for video transcoding and content delivery.Implemented large portions of first generation Web site backend, including role-based security model in PHP using Kohana framework.Advised on projects, technology initiatives, and staffing needs.

Greenwich, Ct, Us

Spent time in consulting, in healthcare, at the Sony Music incubator, and at various early stage startups: along for the ride on 2 minor IPOs. Built a lot of stuff. Learned a lot of stuff.