Detection Analytics Software Engineer

* Description
* ** 5 Month Project with chance of extension, but project shouldn't take more than 8 months.
Remote candidates will be accepted, but candidates THAT CAN COME ONSITE 3 DAYS/WK ARE PREFERRED!
IMPORTANT, PLEASE READ:
This program is made up of 2 projects, with one team on each. Each project team will have 2 developers + 1 PM + 1 Data Scientist (and possibly an ME/EE SME). Two separate teams will execute in parallel (retrofit HRU vs new-build CDU) with minimal crossover due to scope, timeline, and deployment constraints. This role will pair with a Platform Integration Role
TEAM OVERVIEW:
The Signals Quality team in Microsoft's CO+I IDEA group deploys high-reliability detections across data-center telemetry. The team blends analytics (Python/KQL), cloud-native engineering (Azure Functions/Logic Apps), and CI/CD hardening via Azure Dev Ops. Close collaboration with Solutions Engineering enables rapid low/no-code deployments that later graduate into production-grade components.
PROJECT OVERVIEW:
This role focuses on translating validated detection logic into deployable analytics for two environments:

* Retrofit sites using HRU (Heat Rejection Units) and

* New build sites using CDU (Cooling Distribution Units).
Work streams are dedicated per site to meet aggressive timelines. Deployments target a Phoenix-area site and a Milwaukee-area site; no travel is expected.
THIS ROLE'S CONTRIBUTION:
You will design, implement, validate, and tune time-series detections over large telemetry sets. You'll author advanced KQL (ADX) for data shaping and replay, build Python-based analytics (often in notebooks) to iterate quickly, and package logic into Azure Functions/Logic Apps with CI/CD in Azure Dev Ops so detections can be rolled out reliably at scale.
TOP

SKILLS:

1) Advanced KQL / Azure Data Explorer (ADX)

* Author time-series queries (e.g., make-series, make-list, joins, windows) to extract, aggregate, and diagnose telemetry at scale.

* Build replay datasets and run historical backtests to validate threshold/anomaly logic and severity classification before production.

* Identify data gaps and propose telemetry enhancements, map operational scenarios to available signals.
2) Python for Analytics (incl. notebooks) - 7+ YOE

* Wrangle/transpose large datasets; implement feature engineering and transformations needed by detection logic.

* Prototype and harden anomaly/threshold detection logic; collaborate with DS to quantify precision/recall tradeoffs and alert fatigue reduction.

* Package analytics into Azure Functions or glue code around Logic Apps when appropriate.
3) Azure Functions / Logic Apps + Azure Dev Ops CI/CD - 3+ YOE

* Convert proven analytics into deployable components; manage pipelines, artifacts, secrets, and approvals in ADO.

* Integrate detection outputs into downstream systems and operator-facing surfaces; contribute to runbooks and validation plans.

* Partner with Solutions Engineering to deliver rapid low/no-code wins, then graduate those into hardened CI/CD-backed services.
DAY TO DAY RESPONSIBILITIES:
- Analyze large telemetry sets; shape/replay datasets; validate detection effectiveness and tune severity to reduce alert fatigue.

* Translate prototype logic into production-ready analytics (Python/KQL); package into Functions/Logic Apps and wire up CI/CD.

* Work with Solutions Engineering on low/no-code accelerators; coordinate with Engineering to harden and scale solutions.

* Validate data paths and telemetry onboarding; triage signal gaps with domain SMEs (CDU/HRU).

* Track work in ADO boards; contribute to design notes, detection descriptions, validation plans, and weekly status for the PM.
BEST VS AVERAGE:
Best:
- Moves fluently between notebooks and ADX
- Quickly builds/replays time-series detections
- Collaborates cross-functionally to land Logic Apps/Functions with robust CI/CD
- Communicates clearly about tradeoffs and outcomes.

* Average:
Strong coder but slow on time-series KQL or dataset shaping, relies heavily on others to product ionize logic; struggles to tell the story of results for operators and PMs.

* Skills
* c-sharp, python, azure, ci/cd, Kusto, KQL, powershell, github, devops, machine learning, AI

* Top Skills Details
* c-sharp,python,azure,ci/cd,Kusto,KQL,powershell,github,devops

* Additional

Skills & Qualifications
* N/A

* Experience Level
* Expert Level
We reserve the right to pay above or below the posted wage based on factors unrelated to sex, race, or any other protected
classification. Eligibility requirements apply to some benefits
and may depend on your job classification and length of employment. Benefits are subject to change and may be subject to specific elections, plan, or program terms. This temporary role may be eligible for the following:

* Medical, dental & vision

* 401(k)/Roth

* Insurance (Basic/Supplemental Life & AD&D)

* Short and long-term disability

* Health and Dependent Care Spending Accounts (HAS & DCFSA)

* Transportation benefits

* Employee Assistance Program

* Time off/Leave (PTO, Vacation, or Sick…