Inside Darwin Analytics

A small look into the development side of Darwin Analytics.

Notes: https://gist.github.com/jelmersnoeck/5005575

Transcript

1. MORE INSIGHTS, LESS METRICS.
2. The tree has all the answers.
3. data sources
4. data sources magic
5. data sources magic profit++
6. Overview & insights Advice Future
7. Overview & insights Advice Future
8. Overview & insights Advice Future
9. Answers to what?
10. Which channel is most effective in reaching your marketing objectives? Effect
11. Which channel is most cost-efficient in reaching your marketing objectives? Cost
12. How am I evolving towards my marketing objectives? Evolution
13. Answers for who?
14. Mark Eaters
16. Third party connections data sources
17. Business logic layer magic
18. Harvesters Business logic layer Processors
19. Business logic layer Processors Harvesters
20. Business logic layer Processors Harvesters
21. Business logic layer Processors Harvesters
22. Business logic layer Processors Harvesters
23. Business logic layer Processors Harvesters
24. Presentation layer profit
25. Future: API layer
26. “#TDD is the process of describing what code should do before writing it. Saying that you can’t do that is saying that you can’t think!” ~ @everzet
27. Fail fast, succeed faster. 130+ tests, 350+ assertions Still too low (no functional tests)
28. “Ink is better than memory” phpDocumentor 2 16k+ ncloc 8k+ cloc
29. Vagrant
30. Vagrant dependencies
31. Vagrant dependencies easy configuration
32. Parallelization
33. Parallelization Multiprocessing
34. Parallelization Multiprocessing Streaming
35. Parallelization Multiprocessing Streaming Message Queue
36. Machine learning
37. Build server
38. Whoops, there’s my exit! @jelmersnoeck