Job Description:
SQL Data and Statistical Analyst
Support the team that tests autonomous vehicles.
Core responsibilities:
? Create SQL queries to pull and aggregate data from tables of logs of autonomous driving
from the real world and simulation.
? Correctly estimate rates of different types of behaviors with confidence intervals from
sampled events using SQL. Experience with confidence interval calculation is required.
? Create dashboards to display rate estimates and simulation results with appropriate filtering
capabilities.
? Follow-up with data scientists, systems engineers, and test engineers.
? Update queries for existing dashboards to reflect the latest release.
? Present overall status of rates and specific contributing logs and simulations.
? Improve the workflows by automation.
Required:
? Strong ability to write SQL queries, SQL modules, and SQL tests, and maintain a high
standard of readability and code health.
? Strong knowledge of statistics, including confidence interval calculations.
? 3-5 years of data analysis experience.
? Previous experience in program management, task tracking, or operations.
? Previous experience with program management software tools. (JIRA or equivalent)
? Strong understanding of databases and data pipelines.
? BS/MS in a technical field.
Desired:
? Experience in automotive and/or trucking industry.
? Experience implementing new operational workflows.
? Interest and willingness to improve existing analysis software and report generation.
? Python and/or C++ programming experience.
Service deliverables
? Interaction model
? Analysts will be assigned task in our internal tracking tool (buganizer) with priority,
due date, and stakeholder specified. Alerts of new tasks will be emailed.
? Analyst will attend weekly meetings, report on progress and results, and alter or
reprioritize tasks as discussed.
? What is success?
? Rate estimates are automatically and correctly calculated.
? Quality of the rate estimates is monitored continuously. New estimates are
calculated correctly the first time.
? Dashboards provide needed functionality for key stakeholders to understand rates
and the contributing events quickly and easily.
? Updates to tables and dashboards are completed within deadlines; slippage is
communicated promptly.
? All code is code reviewed and submitted to our code base with established
processes.
? All SQL code is unit tested and end-to-end tested.
? SQL code adheres to go/sqlstyle (viewable copy at
SQL Style Guide - SQL Readability.pdf ) with minimal review comments.
? What metrics will be used to measure success and performance?
? Latency in adding new topics for rate estimation.
? Number of errors in rate estimates that are escalated by outside stakeholders.
? Number of review comments in change lists submitted to the code base. Fewer is
better.
