TNT-7956: add applied-ai-pr-automated.md

PULL_REQUEST_TEMPLATE/applied-ai-automation/automated-checks.md

@@ -0,0 +1,11 @@
+# 🤖 Automated checks
+
+---
+
+#### ❗️ AUTOMATICALLY GENERATED READ-ONLY SECTION❗️
+
+---
+
+<!-- automated-checks-start -->
+> ⏳ *Automated checks are pending. GitHub Actions will update this section shortly.*
+<!-- automated-checks-end -->

PULL_REQUEST_TEMPLATE/applied-ai-automation/manual-checks.md

@@ -0,0 +1,25 @@
+
+# 🧑‍💻 Manual checks
+
+---
+
+#### ❗️THIS SECTION MUST BE COMPLETED BEFORE PR APPROVAL❗️
+
+---
+
+*Confirm ALL checks:*
+
+- [ ] 🔙 I can rollback this change if something goes wrong
+- [ ] 👓 I will monitor the live service after deploying this change or have arranged for someone to do so
+- [ ] 📄 I have updated documentation or this is not applicable
+
+*Select ONE 🐈‍⬛ of the following 3 options:*
+
+- [ ] 🐈‍⬛ Shadow deployment **is available** in the "Automated checks" section
+- [ ] 🐈‍⬛ Shadow deployment **is not available**, and I have tested that my code changes work in a local or staging environment. Link: `<insert link to locally triggered test run or description of local testing>`
+- [ ] 🐈‍⬛ This is a CI/CD change, and is not subject to shadow deployment or local testing, but I have tested the changes before merging to main.
+
+*Select ONE ⌛ of the following 2 options:*
+
+- [ ] ⌛ I confirm that I am deploying this change in the morning / early afternoon and not on a Friday or during code freeze
+- [ ] ⌛ I have obtained approval from my function lead to deploy this change outside of the regular hours

PULL_REQUEST_TEMPLATE/applied-ai-automation/model-checks.md

@@ -0,0 +1,20 @@
+
+# 🧑‍💻 Model checks
+
+---
+
+#### ❗️THIS SECTION MUST BE COMPLETED BEFORE PR APPROVAL❗️
+
+> *Shows/hides AUTOMATICALLY depending if model files have changed*
+
+---
+
+*Confirm ALL checks:*
+
+- [ ] 📈 I have confirmed the performance of the model is as expected (see "Automated checks" section for shadow training pipeline run)
+- [ ] ⁉️ My changes do not introduce a discrepancy between training and inference (e.g. features used, data transformations, etc.) 
+  - check training/inference consistency with Copilot review by using `/compare-training-inference` as a PR comment
+- [ ] 🐈‍⬛ I ran inference using shadow deployment or local testing and confirmed the model is working as expected (see "Automated checks" section for shadow inference pipeline run)
+- [ ] 🕵️ F&I have validated and signed off on the model performance and monitoring plan
+- [ ] 👍 I have received approval from my function lead to deploy this model
+- [ ] 📊 New circuit breaker thresholds have been communicated to Engineering, or I have confirmed that no changes are necessary

PULL_REQUEST_TEMPLATE/applied-ai-automation/pr-description.md

@@ -0,0 +1,3 @@
+# 📝 PR Description
+
+> Replace with short **MANDATORY** description on **WHAT** this change does and **WHY** it is needed

PULL_REQUEST_TEMPLATE/applied-ai-pr-automated.md

@@ -0,0 +1,41 @@
+# 📝 PR Documentation
+*Description*
+`<In 1-2 sentences, describe what this change does and why it is needed.>`
+
+## ✅ Release Checklist
+THE RELEASE CHECKLIST MUST BE COMPLETED BEFORE THIS PR IS APPROVED.
+=========================
+
+
+### 🤖 Automated checks
+
+<!-- automated-checks-start -->
+> ⏳ *Automated checks are pending. GitHub Actions will update this section shortly.*
+<!-- automated-checks-end -->
+
+### 🧑‍💻 Manual checks
+
+  - [ ] 🔙 I can rollback this change if something goes wrong
+  - [ ] 👓 I will monitor the live service after deploying this change or have arranged for someone to do so
+  - [ ] 📄 I have updated documentation or this is not applicable
+
+*Select ONE 🐈‍⬛ of the following 3 options:*
+- [ ] 🐈‍⬛ Shadow deployment **is available** in the [Automated checks](#automated-checks) section
+- [ ] 🐈‍⬛ Shadow deployment **is not available**, and I have tested that my code changes work in a local or staging environment. Link: `<insert link to locally triggered test run or description of local testing>`
+- [ ] 🐈‍⬛ This is a CI/CD change, and is not subject to shadow deployment or local testing, but I have tested the changes before merging to main.
+
+*Select ONE ⌛ of the following 2 options:*
+- [ ] ⌛ I confirm that I am deploying this change in the morning / early afternoon and not on a Friday or during code freeze
+- [ ] ⌛ I have obtained approval from my function lead to deploy this change outside of the regular hours
+
+
+<details id="model-changes-only">
+ <summary>Model changes checks (opens automatically)</summary>
+
+  - [ ] 📈 I have confirmed the performance of the model is as expected (see [Automated checks](#automated-checks) section for shadow training pipeline run)
+  - [ ] ⁉️ My changes do not introduce a discrepancy between training and inference (e.g. features used, data transformations, etc.)
+  - [ ] 🐈‍⬛ I ran inference using shadow deployment or local testing and confirmed the model is working as expected (see [Automated checks](#automated-checks) section for shadow inference pipeline run)
+  - [ ] 🕵️ F&I have validated and signed off on the model performance and monitoring plan
+  - [ ] 👍 I have received approval from my function lead to deploy this model
+  - [ ] 📊 New circuit breaker thresholds have been communicated to Engineering, or I have confirmed that no changes are necessary
+</details>