Enterprise AI Digest #21 - PODCAST on Spotify and Apple
Topics
D365 Managed Operations
Copilot Studio Quotas and Limits
Microsoft Fabric Lakehouse
Events
D365 Managed Operations
Organizations increasingly rely on the platform for global-scale operations, the need for Managed Operations has emerged as a cornerstone for ensuring stability, reliability, and resilience.
What is Managed Operations?
Managed Operations is a suite of features designed to help organizations maintain operational excellence through:
Change Management: Avoid unexpected disruptions and maintain application reliability.
Advanced Insights and Monitoring: Detect and resolve performance issues proactively.
Resilient Data Management: Protect and recover data effortlessly, even in the face of challenges.
Core Capabilities of Managed Operations
Controlled Deployments for Stability
Integrated DevOps Tools: Enable admin-approved pipelines for safe, efficient deployments.
Enhanced Testing: Use Power Apps Test Engine (preview) to automate testing, simplify workflows with YAML-based Power Fx tests, and ensure high-quality releases.
Deployment Visibility: Monitor deployments in real-time with the Maker Portal and Power Platform Admin Center, ensuring approvals and failures are easily managed.
Advanced Monitoring and Insights
Built-In Metrics: Get out-of-box insights into app health and operational performance, allowing rapid detection and triage of issues.
Integrated Data Views: With Fabric Shortcuts (private preview), connect Dataverse to Fabric for instant access to up-to-date Power Platform and Dynamics 365 logs, enhancing decision-making and streamlining data management.
Recommendations in Action: Proactive guidance helps optimize app performance and reliability.
Unmatched Data Resilience
Extended Backup Retention: Retain backups for up to 28 days, offering more flexibility for recovery.
Self-Serve Disaster Recovery: Private preview of disaster recovery tools lets admins perform drills, failovers, and failbacks seamlessly.
Recycle Bin for Deleted Records: Safeguard against accidental data loss with configurable recovery windows for deleted records.
Long-Term Data Retention: Archive data securely in Azure Data Lake tiers for compliance and cost-efficiency.
For more details, refer Microsoft blog.
Copilot Studio Quotas and Limits
Microsoft Copilot Studio is a powerful tool for building and deploying AI-driven agents, but it's essential to understand the quotas and operational details to maximize its potential effectively.
Message Quotas:
Copilot Studio enforces limits to maintain stability. Paid plans allow up to 8000 requests per minute (RPM) per Dataverse environment. If the limit is exceeded, users will see a "Send failed. Retry." message, ensuring system resilience against unexpected usage surges.
Operational Limits:
Limits for features like topics, agents, and skills depend on whether you're using the web app or the Teams app. For example, a web app supports up to 1000 topics per agent, while Teams apps are capped at 250 topics per agent unless upgraded. In scenarios like Omnichannel integrations, message size limits of 28 KB can impact transfers, especially when passing large sets of variables.
Subscription Details:
Standard subscriptions include 250,000 Power Platform requests every 24 hours, while Teams plans support 6,000 requests per day. If more capacity is needed, additional sessions or Power Platform requests can be purchased.
App Registration and Security:
Copilot Studio automatically manages app registrations, ensuring secure communication with channels and skills. Certificates are rotated regularly to comply with security standards, requiring no manual action from users.
Required Services:
For uninterrupted functionality, specific domains like .powerva.microsoft.com and .directline.botframework.com must be accessible. Proper configuration ensures seamless operations for bot interactions and analytics reporting.
Understanding and managing quotas, limits, and required configurations ensures Copilot Studio agents deliver maximum value with minimal disruption.
Microsoft Fabric Lakehouse
Microsoft Fabric is an all-in-one analytics platform that integrates data movement, engineering, science, real-time analytics, and business intelligence.
Building a Lakehouse with Fabric Fabric allows organizations to implement data lakehouse architectures like the medallion approach, organizing data into Bronze, Silver, and Gold layers for raw, validated, and refined datasets, respectively.
Key Features:
Unified Data Storage: Fabric standardizes on the Delta Lake format, allowing seamless access and manipulation of datasets across engines.
Flexible Architecture: Implement medallion architecture or data mesh models based on your organizational needs, ensuring scalability and efficiency.
Simplified Data Transformation: Choose between low-code/no-code pipelines or advanced coding with Notebooks to suit your team's skill set.
Powerful Consumption Capabilities: Power BI integrates directly with the lakehouse for dynamic reporting and dashboards. SQL endpoints facilitate easy connectivity with other analytics tools.
End-to-End Workflow Example Using the Wide World Importers (WWI) sample data, this process demonstrates ingesting historical and incremental data into the lakehouse, organizing it into layers, and analyzing it through Power BI.