Wikimedia Discovery/Team/Analyst onboarding

Jump to navigation Jump to search

Setting computers up, setting accounts up and getting known[edit]

  • Read Wikimedia Apps/Team/Onboarding/Various wikis of Wikimedia, and how to log in to them which might be helpful with some of the later steps described here
  • Generate two SSH keys (see this article on GitHub for instructions): one for our analytics machines and one for labs ("Wikitech") access.
  • Create a Wikitech account. Go to Preferences → OpenStack and include the public part of a newly-generated SSH keys, and add both public keys to your Gerrit account. Get Mikhail to give you admin access to the relevant projects (discovery-stats and shiny-r).
  • Create a Phabricator account by logging in with your MediaWiki account (akin to signing in via Facebook or Google). Get Mikhail to sign you up to all the appropriate boards (e.g. Analysis Sprint Board and Discovery Board) and walk you through the process of working with cards.
  • File a Phabricator ticket in the Ops-Access queue, CC-ing Dan and Mikhail, containing the public part of a newly-generated SSH key, asking for access to stat1002 ("stat2"), stat1003 ("stat3"), and fluorine. Don't worry about what those names mean just yet.
  • Get an IRC client. Join #wikimedia-discovery connect. Welcome to our team live chat!
    • Learn about cloaks and request one. (It can take a few days for the cloak stuff to be resolved.)
    • Additional channels to join: /join: #wikimedia-discovery, #wikimedia-analytics, #wikimedia-research, #wikimedia-ai, #wikimedia-operations
  • Schedule staff photo (just list your name here)
    • Or you can list yourself as not wanting to have your photo up by adding your name under the cloak of invisibility list.
  • Add yourself to the team page—twice: in the info box at the top of the page, and under the section "The Team | Members".
  • Subscribe to some mailing lists using your account
  • Setup your SSH configuration (refer to the EOM supplement on Office wiki). Check Saving SSH keys in macOS Sierra keychain.
  • In Terminal on OS X, go to Preferences → Profiles → Shell. Check Run command and type in: eval ssh-agent and check Run inside shell.
  • Possibly get Mikhail to poke Andrew Otto (ottomata on IRC/Phabricator) to add you to analytics-search-users group.
  • Poke Deb to introduce you on wmfall (so all WMF staff are aware of you joining us because you're awesome and it's so great that you're on our team now).

Meeting interesting people[edit]

This list should take about 2 weeks to get through. A thing to keep in mind is that people are bound to mention technical things that they're working on or have worked on that you may not have familiarity with. Don't be afraid to ask "what's that?" We're not gonna fire you because you had to ask what a web request is or what JSON is :P

  • You should sit down with Deb Tankersley – Discovery's Product Manager – and discuss the department's goals and get an overview of the various teams and how Analytics supports the team's work.
  • You should sit down with Erika Bjune – Engineering Director of the Search Platform – and discuss how Analytics supports the backend Search team's work.
  • Grab Trey Jones, a wicked-smaht text processing specialist, linguist, and our relevance engineer and he knows mathz too
  • Grab Erik Bernhardson, who knows far too much about our search systems and event logging.
  • Talk to Chris Koerner (Discovery's Community Liaison) about the community!
  • Talk to Aaron Halfaker about research in the context of Wikipedia users, and Ellery Wulczyn to chat about data analysis at the Foundation.
  • Talk to Stephen LaPorte from legal about privacy and data sources.
  • Talk to Helen Jiang who is a data analyst in Editing.
  • Make sure Mikhail has added you to the research group meetings.
  • Attend the Community Liaison Cabal meeting to ask questions and hear about the way people interact with our community.

Working on projects[edit]

  • Go through the Dashboarding & Testing presentation Oliver and Mikhail gave at All Hands 2016. Make a list of questions. Try to answer as many of them as you can by reading through our Employee Operations Manual on Meta. This is less a test for you and more of a test for the thoroughness of the EOM :) Mikhail will answer whatever questions you can't find an answer to.
  • Check out the statistical graph literacy talk Mikhail gave (YouTube recording, Slides)
  • See if you can run this presentation written in RMarkdown.
  • Pick a small (low point value) task from Phabricator. Work on it! Reach out to Mikhail if you run into any issues or it requires specific domain knowledge.
  • Talk through big upcoming projects with Deb and Mikhail and how you fit into them. Deb will setup a meeting after your first couple of weeks to see how you're doing.
  • Check and update this page in the first week or two if your experience was different than how it was described here to be. Your first major task is to spot holes in this thing we call a wiki page, fill 'em up, and make onboarding better for the next person. :)

Notes for supervisor and/or co-analyst helping with onboarding[edit]

  • Refer to T142648 for example of requesting access.
  • Refer to T142436 for example of how to structure the process of learning to work with dashboards.
  • Refer to T143128 for example of how to structure the [EPIC] process of learning about our databases: