Context

At OptimalBI we roadmap by breaking “things we want to do” in to three categories:

  • “Arm waves” are broad promises about our focus areas for a product
  • “Hand waves” are core capabilities and value propositions needed to deliver an arm wave
  • “Finger points” are user stories required to deliver a hand wave over a sprint or three

Listed below are the “arm waves” for the Optimal Data Engine: the things we’d love it to do.

When we work out the “hand waves” that deliver each arm wave, we fill them in here so you know what we’re working on. Greyed-out text means “we know we need this but haven’t delivered it yet”, black text means “Done”.

Finger Points

Things we are actively working on right now.

  • Documenting the current set of features
  • Setting up Test Driven Development and  Continuous Integration capabilities
  • Creating reusable tests for each feature
  • 3CpO Oracle ODI version

Continuous integration

We’ve decided to follow the best of modern software development practice to build the Optimal Data Engine. These techniques are quite complicated for data warehouse development, but the payoffs are huge.

Hand waves we are working on:

Backlogged Finger Points

Things we are actively thinking about working on right now.

  • Point in Time

Hand Waves

Continuous integration

We’ve decided to follow the best of modern software development practice to build the Optimal Data Engine. These techniques are quite complicated for data warehouse development, but the payoffs are huge.

 

Hand waves we are not yet working on:

  • Automation of absolutely everything
  • Capturing runtimes of builds and tests

Business rules

We’ve seen business rules kill many data warehouse projects; getting them right is notoriously difficult. The minimum components (hand waves) we think we need for success are:

  • A config-driven rules engine
  • A user interface to that engine
  • A standard way of capturing the outputs after rules are applied, so these are governed like any other data in ODE

Onboarding Developers

We intend the Optimal Data Engine to be a community project, and we’re looking for partners to join us in developing it!

To do this we need to deliver the following hand waves:

  • Training in how we use Git for version control
  • Onboarding so that trained contributors know how to use our continuous integration environment
  • Unit testing cases and tool recommendations to help collaborators write better code

Virtualised Star Schemas

Creation of virtualised Dimensional and Facts tables via the use of ODE configuration.

Arm Waves

Logging and monitoring

  • Logging orchestration runtimes

What we have delivered

Features

We’ve built some features, which you can read here.

Standards

We’ve established some standards, which you can read here.