Skip to content
English
  • There are no suggestions because the search field is empty.

v20251203

Release notes

This release contains improvements to the initial release of TwinThread Advisor as well as other improvements in the recently released Perfect Centerline model.  It also contains many issue corrections.

Issues

82078 - ML Pipeline: format annotation tables in gfm format

84389 - Save data manifest report even when aborting for missing data

82433 - Helm Chart Dependencies Upgrade

81921 - Fix Content Duplicate Error

83202 - Dynamic Table: Fix Download Button Positioning

82212 - Lock org deploy ops behind more strict access key

82485 - Cleanup ModelTaskContent in more AssetModelId deletion references

82495 - Model pipeline hotfixes against main build using TT common dev

82527 - Error boundary on notebook nodes

83110 - Anti-forgery token does not refresh after 2 hours - 1 minute due to cookie still existing

82669 - Turn on advisor collection

82461 - Ability to trend multiple mapped sensors from the datasource sensor list

83872 - Throughput Optimization Model stopped producing Predicted Property

82216 - Markdown Content: Model Status

82224 - Model Quickstart Page: redesign layout to allow space for model objectives & link to model status

84394 - Model Status - link to property detail

84356 - Calculate summary statistics OOM on evaluate

82489 - MOM failed to train with error Can only use .str accessor with string values

82338 - Anomaly Model training never completes

81630 - PCL: topics wizard item for distinct event defining production runs separate from segmentation

82880 - Notebooks associated with a model should be deleted when the model is deleted

84355 - Create Evaluate codepath for PCL and Detect Anomaly

84379 - Configure Model modal overwrites user-provided text every ~1-2sec

82612 - Change org deploy status to take query string params

82229 - Infrastructure to generate notebooks automatically on models

81928 - PCL Report: Handle different specified transform for objectives

81774 - LLM: knowledge base scraping enhancements

82365 - Create Generic Document CRUD

83899 - Pass Stale Quickstart Tabs to the frontend

82299 - Fix Perfect Centerline Performance

82794 - CORS Misconfiguration: Arbitrary Subdomains Trusted

82846 - Add enablement record to tt-master

84274 - fix internal api query controller

83053 - Add DataDog logging for Advisor

83458 - Add LLM Orchestrator to Serverside Solution

83769 - Simple Batch Upgrade

30461 - Add next run time to trigger tiles

82827 - Move notebook owner auth logic to stored procedures

84210 - Frontend route to resolve org switching and ad hoc content creation

80678 - Not able to change the organization of some analysis

84392 - Use Simple Table format for lifecycle in Model Status

82139 - Fix TwinThread-Global build warning

83428 - Model Manager Agent: create llm background tasks during model lifecycle

83837 - Min Max Toggle FOR PCL Report

83947 - Assess Performance Issues Parent Ticket

82981 - Advisor: Add user policy level authorization in Advisor Router

77361 - Event Based Model Failing

82698 - Dependabot security alert

81109 - Timeline v2 update

82871 - The very bottom of the markdown node seems to be getting cut off

81898 - Notebook - when user is tagged and receives email, direct user to the exact location where of the tag that triggered the email

82213 - Advisor: generate python code based on user prompt

83008 - Advisor Notebook context improvements

83091 - Chasing prometheus and Grafana Upgrade issues

84252 - Fully Remove Advisor Feature Flag Checks

82207 - Kde plots on Centerline model is odd

84444 - Add an edit button to the overview section that opens up "configure modal"

81714 - Engine: don't swallow BinaryDeserialize OOM exceptions

82941 - Change colors on PC report graphic mode to make target and current values more visible

83421 - Recommendation Rejection

41340 - Anomaly History safe range stated is not clear compared to the normal distribution

83856 - Segmentation Topic Wizard steps: removing these topics should unmap vs delete underlying topic

84036 - Define if Documents are Editable on the DocumentType, not Document itself

81274 - Missing Events when Backfilling from CDS

53570 - Fix implicit vulnerabilities identified on ML Pipeline container images

83703 - Turn on PCL collection

83746 - Address CS Team advisor data manifest expected query answers

82608 - Remove Event Distribution legend from anomaly model KDE plots

82632 - K8 Cluster Upgrades

82142 - Predict Property model failing execute when quality objective is missing

83426 - WorkExecutorJob => AI Foundry Secrets

84367 - Remove Data Manifest generation from Train blocks to avoid duplication with OF

82165 - Property Input Search In Add Data Wizard not working

83741 - Model Artifact List Endpoint

83976 - Let internal api TT NEU select only orgs in tt-neu

83530 - Ability to hide/show min and max centerlines for Best and Recent runs on centerline report

83948 - Metric thumbnail trends not drawing properly (should be time period of content)

84419 - Do not generate Model Status When Advisor is not on

75579 - MOM run history not working

82215 - Markdown Content: Model Objectives

84390 - Artifacts page: have content drilldowns have breadcrumbs back to artifact page with content id focuses

83482 - Perfect Centerline - Sum Objective Handling

81936 - LLM: use time select for interpret content

83280 - Transitioning to full-screen mid-response severs Advisor connection

82920 - [FEAT] Automatic WebApi Build and Deploy Pipeline Trigger

83163 - Invalid Markdown Table formatting

82422 - Advisor: tool-calling for model lifecycle chronology

81926 - Fix spModel_GetPropertyMap

84352 - Provide Internal WebApi endpoint for toggling capabilities for Testing

84165 - Model Lifecycle: add prefix and workflow errors

83117 - Can we add exception for the notebook endpoints to the WAF

84044 - Advisor: Evaluate Summary

84285 - Remove redundant status bar in model factory

83757 - Run Performance - Centerline

84429 - Model Quickstart Next Step: have "nudge" to review Evaluation Summary Report when it is available

83603 - Dynamic-code-service Restart Job Failure

82017 - Import Model Task Annotations into the Database

82855 - Setup LLM process to generate Model Status Report

82541 - S101 Ideal Running Conditions

83859 - Training Failure

83418 - Capability Enablement: Utilize Tenant Orgs to store capabilities

83155 - 500 error when trying to create an adhoc timeline on a specific twin

73020 - Update Adjust Recommendations template thread to check for null conditions in Adjust Recommendations block

83106 - Advisor - User Exception, don't show as error to user, just return as regular message

78900 - datadog messages missing "OrgKey"

83423 - Model Training Error - Maximize Throughput

83052 - Process Generic Model Task Markdown files

73690 - Audit existing Evaluate Content and pre-flight checks: what are we missing?

83055 - Special MES Events to not seem to map correctly

83034 - Advisor: Add example prompts for notebooks

82419 - SP: model lifecycle chronology

82458 - Advisor: static python dataset analysis script

81572 - PCL: topic wizard should be extended to allow definition of Production Runs

84463 - Various Advisor Fixes

82367 - Recommendations being filtered when its unexpected

81925 - Search option in Add Data wizard is not working for inputs and controllables section

82847 - Change feature flags for enablement on frontend

81767 - PCL: pass overrides for user-created segments from overrides screen

79279 - Engine: replace deprecated BinarySerializer

82849 - Return feature enablement in org open

82437 - Engine: some CreateDataset[Batch] models failing due to unicode chars

81947 - Anomaly History: missing placeholders

81830 - Allow Perfect Centerline Report to swap segments live

82797 - Create content type that crashes intentionally

83722 - Create Model Artifacts

82115 - Use TwinThread.Global version of Config class in ServerSide repo

82220 - Create Model Manager Agent

80526 - We need to automatically suspend any Time or Event based model triggers in the specific model when EOTR is kicked off.

82394 - In run recommendations not enable-able is confusing

82479 - Add additional Org Metrics to better understand customer usage

82675 - Ideal mode of operation - PCL

82791 - Insecure MIME Type Configuration

82859 - Fix spark operator deployments

83188 - Power users w/o "class" policy can still create, delete classes but cannot view them

51652 - Predict & Optimize Property: create production runs outputs

83719 - Delete model notebooks list table

83246 - New custom dataset types were not letting cs team members create instances of them in prod orgs

83282 - MOM Model failing to train

82788 - Secrets Checked Into Github

82768 - Production Runs Box and Whisker: fetch data from train tasks

80801 - MoM Recommendation Filtering by Action Threshold

81640 - Edit topic modal can get out of sync in the topics wizard

83349 - CPK Not populating on Run perf

83606 - Advisor Model Support: make data manifest files good

81093 - Evaluate: start/end offset recommendations confusion

83503 - Advisor Model Support: Suggested Next Step

82475 - lookback trigger time not visible in-app

82019 - Operationalize page not showing Quickstart execute tile error

81035 - LLM: Review error handling and reporting - setup dashboard to easily view all errors

83033 - SE Compass weather data

79427 - Historize output of spSubscription_UsageReport

80732 - LLM: generate help article about TT Advisor

77962 - Twin Images Not Honoring Global Time Selector

82704 - Run Performance Analysis

82911 - Copilot instructions - set apply to webapi directory to filter included instructions

84251 - Model "Quick Start" should be "QuickStart"

83892 - MoM model Objectives not getting assigned when Class is applied to Twins

82175 - Ability to search for properties under multiple twins for Timeline V2

82021 - Recalcs alternating between running slow and fast over each one day interval

83798 - CreateNewOrg - Add secondary Azure connection

83838 - Add Model Evaluation Summary Report md to Evaluate Page if it exists

81631 - PCL: back-end to support a distinct event configuration to define production runs

82159 - Fix tt-master dodatasync

82077 - Evaluate Operation: default to train time range, with a warning if this would result in a long run-time

81913 - Centerline Report is not showing the manually overwritten centerlines

82830 - LLM Advisor: Expand NotebookId into associated context

82336 - Model quickstart status flickering

82713 - Need to clean up model status reporting for Class Models

83607 - Perfect Batch Model: set min feature score param to 0 by default

82702 - Cleanup Guided analytics dev

81696 - LLM: export prompts and review

83420 - Perfect Centerline Model Train Failure

84468 - EDF/Batch integration

83124 - Create New Org - Regularly refresh WorkerAPI access token

83229 - Calculate Summary Statistics: Failsafe to only process each property only once

82217 - Markdown Content: Model Task Annotations

83290 - Invalid ideal conditions due to incorrect training data values

83515 - New Org Deploy: Validate Organization Name

82772 - Fix sshd_config on model pipeline

83987 - Implement 10 minute Cache Invalidation for Capabilities

82219 - Create Model Status Notebook Template

82226 - Create Notebooks list page on any model

81878 - Save anomaly history by chunk by default

81977 - Update DynamicCodeService proto file field names

80359 - PC: handle un-modeled segment overrides at Execute time

84231 - e2e test models failing to train

82918 - Add policy, and conditionally show Advisor

82868 - When a PC model is duplicated the objectives are defaulted to Target on the copy model

82829 - MOM: Model select block failing on model sensitivity subtask

82848 - Create API filters for product enablement

82477 - Support Production Run Event in PCL Content

82631 - Need to review user reported Execute Intergration logic that is not working

83027 - PC: Newly added segment and centerlines do not show up under startup recommendations/startup recommendations API

83329 - Use Model Description for Model Objectives

82865 - Perfect Centerlines: on configuring the Production Runs

83476 - Remove Event KDE legend item when there is no event kde to display

83326 - Operationalize page timing rounding issues

78663 - Data Validation: Markdown Content Reports

83966 - Drill downs to to properties should default to at least 30 days (multiples of time period of the assess performance content)

82484 - Model stuck in evaluate

83037 - OPCDA - Make browsing RSLINX tags configurable

83194 - Add data-testId on this "settings" button on

82214 - Advisor: add tool calling to read model objectives and status (markdown)

73138 - Make excel import/export policy based

82140 - Add Model, Type association to Notebooks

82711 - Engine job performance cleanup

84381 - Adjust evaluate data quality (frequency) report

30438 - For Array Properties, use Array Timeline as the default content type

84258 - Poll next step to keep it accurate

78240 - Anomaly not visible at drill in

82030 - Recommendations in opposite direction for few segments

82786 - Improper Permissions on Certificate Files

83538 - KDE + Ideals visual: toggle to show centerlines as well as ideals (when available)

84253 - Webapi exception handling -> log to datadog

81927 - Ensure Workflow errors are fetched immediately in addition to on interval

82106 - Knowledge base soft delete

82870 - Add Notebook context to Advisor

83193 - Power user can walk through " add analysis to notebook/case " until error on submit

84359 - create Evaluate MOM codepath in ML Pipeline

83166 - Production Run Summaries are not displaying on PC

83224 - Better handle CDN asset failure

83050 - Sensor Linking Page Unmap / Delete property button

81378 - Sub-Org Admin user can View parent and other organizations

82510 - Assess Performance Trends Not Rendering Properly

81978 - Re-target all library projects to net8 except those related to on-prem agent

84439 - Show Data Manifest on Eval page when TwinThread Advisor is not enabled

82065 - Create Model Document Viewer

82136 - MOM: model is executing but Get InRun Recommendations thread block is not returning any results

83726 - Clean up model notebook autogeneration

83228 - disable delete button on some model notebooks

81017 - Confusing Eval Message

82402 - Product segment value interpreted as float vs integer/string

81776 - Edit topics modal crashes when clicking event states in the dropdown

82700 - Prep work for doing advisor to work as an executor job

83205 - Advisor - TTApiClient update authentication

82780 - Insecure Bearer Token Generation

83571 - Missing content when drilling down on Historical Anomaly Breakdown from MOM model

82225 - Add Notebooks list page to model Viewing dropdown

83702 - model engine filtering confusion

82795 - All of the agents in the ADO default pool appear to be offline