Enterprise AI Digest #30 - PODCAST on Spotify and Apple
Topics:
D365 Business Central Copilots
Power Platform Wave1 2025
AI: DeepSeek R1 on Azure AI Foundry
Data: Microsoft Fabric Copilots
D365 Business Central Copilots
With Copilot in Business Central, small and medium-sized organizations empower their workers through intuitive AI tools that boost creativity and save time.
Analysis Assist This feature uses natural language to set up analysis tabs in list pages, so you can quickly generate summaries or pivot data (e.g., “sort on quantity from smallest to largest”) without manually rearranging rows, columns, or filters.
Bank Account Reconciliation Assist AI helps match unmatched bank transactions with ledger entries, and if it can’t find exact matches, Copilot suggests the best G/L account (for example, “Fuel Stop 24” might map to a Transportation account).
E-Documents for Vendor Invoices Copilot speeds up mapping vendor invoices to purchase orders, reducing the need to hunt through lines or re-enter data—especially when invoices don’t exactly match POs.
AI-Powered Writing Assistance For product descriptions sold in online shops, Copilot suggests creative marketing text based on the item’s attributes, which you can finalize with minimal edits.
Sales Line Suggestion Copilot helps add product lines on sales documents by finding items (even if the product name and attributes differ) or locating past sales references (e.g., “Need last invoice from April”).
Sales Order Agent This AI-powered tool parses email requests, clarifies any missing details, checks inventory availability, and returns a formatted quote PDF—automatically creating sales orders from multi-turn conversations.
Suggest Item Substitutions When an item is out of stock, Copilot mines item data (like color, category, and barcodes) to propose similar substitutes—helpful for quickly satisfying customer needs.
Number Series Setup Copilot reduces time spent on configuring or updating your numbering systems (e.g., monthly or year-based sequences) by allowing natural language input (like “Set 2025 posted invoices to start from INV-0001”).
Power Platform Wave1 2025
Microsoft has outlined upcoming features for Power Platform to be delivered between April 2025 and September 2025. Delivery timelines and functionality may shift, so keep an eye on official updates. Early access starts February 3, 2025, letting admins turn on certain features before their general availability.
Key Highlights
Application Lifecycle Management (ALM)Deployments Hub: Get pipeline overviews and AI-driven recommendations (Public Preview in Jan 2025, GA in May 2025).
Enterprise-Scale Administration: New Power Platform Admin Center: More robust management UI (GA in May 2025), plus group admin roles to delegate responsibilities.
Security & Compliance: Extended Virtual Network Support, IP Firewall for Copilot Studio, Enhanced Admin Controls for Copilot: Streamline environment access and secure data. Masked Sensitive Fields & Filtered Views: Keep confidential data hidden at the column level and filter out unauthorized users.
DeepSeek R1 on Azure AI Foundry
What Is Azure AI Foundry? Azure AI Foundry is Microsoft’s unified platform for discovering, deploying, and managing a wide range of AI models—from open-source solutions to specialized industry models—all backed by Azure’s enterprise-grade security and performance.
What Is DeepSeek? DeepSeek is an AI model designed to provide powerful, cost-effective intelligence for a variety of enterprise use cases. It specializes in advanced data analysis and natural language tasks, offering state-of-the-art performance without the need for extensive infrastructure.
Why DeepSeek? By balancing high-level AI capabilities with easy integration and cost-efficiency, DeepSeek allows more businesses to benefit from cutting-edge insights. Whether you’re accelerating AI development or looking to enhance existing workflows, DeepSeek’s secure and scalable design—now available through Azure AI Foundry—makes it a strong choice for innovative, production-ready solutions.
How to Use DeepSeek R1 in the Model Catalog
Get an Azure Subscription: If you don’t have one, sign up here.
Search & Open: Find “DeepSeek R1” in the Azure AI Foundry model catalog.
Deploy: Obtain an inference API, key, and direct access to the playground.
Explore: Test prompts in the playground or integrate the API into your workflows (via client libraries or REST calls).
Microsoft Fabric Copilots
Microsoft continues expanding Copilot capabilities across Fabric, offering AI-driven assistance for data professionals, real-time analysis, and user-friendly app creation.
Copilot for Data Science & Data Engineering
Intelligent Code Suggestions: Automates repetitive tasks and provides relevant code templates for building data pipelines and analytical models.
Adaptive to Your Workflow: Whether you’re cleaning data or finalizing insights, Copilot responds to context, speeding up the journey from raw data to finished results.
Copilot for Data Factory
Simple Data Transformations: Generates code to help citizen and pro data wranglers transform data quickly.
Clarify Complex Workflows: Copilot explains the generated code, making advanced data tasks more approachable.
Copilot for Data Warehouse
Natural Language to SQL: Quickly convert everyday language into structured queries.
Smart Code Completion & Insights: Reduces manual query building and offers helpful suggestions for data warehousing tasks.
Copilot for Power BI
Report Creation Made Easy: Generate reports automatically by prompting Copilot on a given topic.
Automated Summaries & Q&A: Summarize new report pages or create synonyms for better natural language queries.
Copilot for Real-Time Intelligence
AI-Assisted KQL Queries: Ask questions in plain English, and Copilot translates them into Kusto Query Language for near-instant analytics.
Both Experts & Beginners: Handy for citizen data scientists or seasoned KQL users wanting faster insights.
Copilot for SQL Database
Streamlined OLTP Tasks: Uses natural language, code suggestions, and Q&A to optimize SQL Database operations.
Document-Based Q&A: Quickly reference relevant docs, ensuring you run the right queries for your OLTP environment.