Download as PDF, PPTX
Uploaded byMLconf
Nikhil Garg, Engineering Manager, Quora at MLconf SF 2016
The document discusses the rationale behind Quora's decision to build their own machine learning (ML) platform instead of buying commercial solutions, highlighting seven key reasons including the need for integration, scalability, and control over ML processes. It emphasizes the blurred lines between ML and product development, the advantages of open-source technology, and the strategic importance of ML as a core competency for Quora. Ultimately, it concludes that a sustainable ML platform is essential for driving innovation at scale.
More Related Content
Similar to Nikhil Garg, Engineering Manager, Quora at MLconf SF 2016
Nikhil Garg, Engineering Manager, Quora at MLconf SF 2016
- 1. Building a MachineLearning Platform at Quora Nikhil Garg @nikhilgarg28 @Quora @MLconf 11/11/16 The Quora Answer To “Build vs Buy” For ML Platforms
- 2. ● At Quorasince 2012 ● Currently leading two ML engineering teams: ○ Content Quality ○ ML Platform A bit about me... @nikhilgarg28
- 4. To Grow AndShare World’s Knowledge
- 7. Over 100 millionmonthly uniques Millions of questions & answers In hundreds of thousands of topics Supported by 80 engineers
- 8. What Slows DownML Innovation?
- 9. ● Pipeline jungles ●Lots of glue code to get data in/out of general purpose packages. ● Strong coupling between business logic, data, ML algorithms and configuration. Curse Of Complexity
- 10. ● Online vsoffline ● Production vs experimentation ● C++ vs Python ● Engineering vs research ● ...even more glue code and pipeline jungles. Clash Of Titans
- 11. ● Hard toreuse existing features, data, algorithms, tooling etc. ● Too costly to even get off the ground. Getting New Applications Off The Ground http://www.qvidian.com/blog/resistance-to-change-sales-organizations
- 12. Many Faces OfChaos
- 13. One ring tobring them all and in the darkness bind them!
- 14. Collection of systemsto sustainably increase the business impact of ML at scale. Machine Learning Platform
- 15. ML Platform: Buildor Buy?
- 16. The Quora Answer:Build For Seven Reasons
- 17. Reason # 7 JustCan’t Buy Everything!
- 18. ● No matterhow powerful the platform is, still need to maintain some form of integration ● This thin integration layer then becomes the platform. ● Real questions -- ○ How much does this in-house layer delegate? ○ How much control does it have over delegation? . Degree Of Integration & Delegation
- 19. Reason # 6 FastScalable Production Systems
- 20. End-To-End Online ProductionSystems ● External platforms at best can deploy “predictive models”, as services, not end-to-end online systems ● Gains come from optimizing the whole pipeline, not just algorithms. ● Latency: tens of milliseconds. Managing sharding, batching, data locality, caching, streaming, stragglers, graceful degradation... ● Real world systems -- boosts, diversity constraints, holes in data, skipping stages, hard filters… sounds familiar? Candidate Generation Feature Extraction Scoring Post Processing Data
- 21. Reason # 5 BlurryLine Between Experimentation & Production
- 22. ● We wantthe same code/systems/tools to work for both experimentation & production. ● But we need to carefully “control” the production code to keep it be fast. ● So need to “control” offline experimentation systems too. Candidate Generation Feature Extraction Scoring Post Processing Data Candidate Generation Feature Extraction Training
- 23. Reason # 4 OpenlyUsing Open Source
- 25. ● Logistic Regression ●Elastic Nets ● Random Forests ● Gradient Boosted Decision Trees ● Matrix Factorization ● (Deep) Neural Networks ● LambdaMart ● Clustering ● Random walk based methods ● Word Embeddings ● LDA ● ... Production ML Algorithms At Quora Candidate Generation Feature Extraction Training/Scoring Post Processing Data
- 26. ● Open sourceis great -- lots of great technologies! ● Commerical ML platforms are also open sourcing stuff. ● Learning and cherry-picking favorite parts from ANY open source systems. ● May write our own algorithms too (e.g QMF) ● Building own platform = controlling the delegation, not lack of delegation
- 27. Reason # 3 CommercialPlatforms’ Offerings Are Not Super Valuable To Us
- 28. ● Main offeringsof external platforms are: ○ Lower operational overhead of running machines ○ Out-of-box distributed training. ● Operational overhead ○ Gets amortized over time ○ Shared with non-ML infrastructure. ● Can often train most models in a single multi-core machine. .
- 29. Reason # 2 BlurryLine Between ML & Product Dev
- 30. ● Answer ranking ●Feed ranking ● Search ranking ● User recommendations ● Topic recommendations ● Duplicate questions ● Email Digest ● Request Answers ● Trending now ● Topic expertise prediction ● Spam, abuse detection ● …. Blurry Line Between ML/Non-ML Product
- 31. Blurry Line BetweenML/Non-ML Data Users Answer s Questio ns Topics Votes Follow Ask Write Cast Have Contain Get Commen ts Get Follow Write Have Have Billions of relationships and words
- 32. Blurry Line BetweenML/Non-ML Codebase ● Integration with other utility libraries/services e.g A/B testing, debug tools, monitoring, alerting, data transfer, ... ● Empowering all product engineers to do ML.
- 33. Reason # 1 MLAs Quora’s Core Competency
- 34. ● ML givesus a strategic competitive advantage. ● Want to control and develop deep expertise in the whole stack. ● Quora has a long term focus -- investment in platform more than pays off in the long term. ● Single most important reason to build ML Platform! ML: Critical For Our Strategic Focus Relevance Quality Demand
- 35.
- 36. ● Anyone doingnon-trivial ML needs an ML platform to sustain innovation at scale. ● Build vs buy decision is not all-or-nothing. ● Surface area and importance of ML are deciding factors in the build vs buy decision.
- 37. Nikhil Garg @nikhilgarg28 Thank You! YES,WE ARE HIRING :)