Lessons Learned from Workday's Search Application Journey from POC to Production
In 2014, the Workday team realized their Lucene-based custom search would not scale or provide world-class relevance for end users. After extensive research and prototyping, Elasticsearch came out on top as a solution.
Three years later, Workday's highest-volume, most critical search use cases — Recruiting - Find Jobs and Find Candidates — are in production on Elasticsearch. With this change, indexing time has gone from more than 10 hours to less than a few minutes. Query results are being served in subseconds instead of timing out at the 20 minute threshold. Also, Workday has gone from lacking relevance in results to leveraging out-of-the-box relevance capabilities in Elasticsearch. Happier end users equals a happier team!
Senior Engineer, Workday
Angela Juang is a Senior Software Engineer on Workday's Search team, where she works on solving challenges in building, scaling, and maintaining Workday's search infrastructure built on Elasticsearch. Over the past 8 years, she has enjoyed learning new skills and technologies while developing web applications at Twitter and building test systems for space research. Angela is Bay Area native and proud alumna of UC Berkeley, where she got her B.A. in Computer Science and Applied Math and B.S. in Business Administration. When not coding or leading a team, she loves doing crafts and spending time with her family.
Head of Search, Data Science & Machine Learning, Workday
Madhura Dudhgaonkar is responsible for leading Workdayâ€™s search, data science and machine learning teams based in San Francisco. Her search organization has spent the last ~3 years re-architecting a decade old custom search engine to provide Elasticsearch based distributed search. Madhura's career spans across SUN Microsystems, Adobe and now Workday. Her experience ranging from being a hands-on engineer to leading large engineering organizations.