Sunday 9:45 AM–10:30 AM in Room 2

Embracing the Monolith in Small Teams: Doubling down on python to move fast without breaking things.

Leon Sasson

Audience level:
Intermediate

Description

Its unlikely for small orgs to have resources to support the fanciest and shiny data tooling out there. Choosing the right tools and tightly coupling them, can can give small, cross-disciplinary teams the power of moving fast, without breaking things. This goes against a lot of common advice, but becomes powerful when applied with disciplines. I'll show you how with Django, IPython, and Postgres.

Abstract

Products are more than code

  • Good architectures gives you flexibility
  • Choosing right tools is a biz decision: why Django
  • Overcoming the operational vs analytical data workload disconnect(age-old OLAP vs OLTP) with Postgres.

Data Products are more than models

  • Coupling your Django models with pandas dataframes: a sharp knife.
  • Seamlessly incorporating external sources with Segment.com
  • The right thing, to the right person, at the right time: smart notifications.
  • Appearing smart with dumb safeguards around noisy data.

Feeding back into the business

  • IPython as a BI tool: how a little python can go a long way.
  • "If you have to copy-paste a dashboard, you're not serious about tracking that"
  • Slicing and dicing customer engagement data
  • What Sales didn't know they could have: engagement data as a lead-gen tool.