From Project Chaos to Instant Executive Updates: How We Connected Azure OpenAI to On-Premises SQL Server

John Koziaris
Chief Executive Officer
How alltasksIT built an AI-powered Project Status Agent that transforms project data into stakeholder-ready updates in seconds.

The Challenge Every Project Manager Faces

Project Managers and Project Coordinators spend an enormous amount of time gathering information from multiple systems, analysing project progress, identifying risks, and preparing updates for clients, managers, and stakeholders.
 
The process is often repetitive:
  • Review project records
  • Analyse ticket activity
  • Check budgets and actual hours
  • Identify overdue tasks
  • Review project milestones
  • Compile management reports
  • Draft status updates for stakeholders
From Project Chaos to Instant Executive Updates: How We Connected Azure OpenAI to On-Premises SQL Server
For organisations managing dozens of projects simultaneously, this can consume several hours each week. More importantly, the information is often already sitting inside business systems but remains difficult to access and interpret efficiently.
 

At alltasksIT, we wanted to solve this challenge using Artificial Intelligence.

The Goal

Our objective was simple: Allow project managers to ask questions in plain English and receive detailed project intelligence instantly.

Instead of manually extracting data from ConnectWise and SQL Server, we wanted users to simply ask:

"Help me provide a project update for my manager including recent activity, project progress and risks."

The Solution

We developed a custom Project Status Agent that combines:

  • Azure OpenAI
  • Microsoft Copilot capabilities
  • On-premises SQL Server data
  • ConnectWise project information
  • Custom business logic

The solution connects directly to SQL views within our on-premises ConnectWise database and uses Azure OpenAI to interpret natural language questions and generate meaningful responses.

The project agent currently consumes information from several SQL views including:

  • ATCS_ProjectsWithTicketTimes
  • ATCS_ProjectPhaseWorkPlan
  • ATCS_ProjectPhases
  • ATCS_ProjectOverview

These views provide visibility into:

  • Active projects
  • Project phases
  • Budgeted hours
  • Actual hours
  • Ticket activity
  • Progress tracking
  • Project timelines

Rather than exposing raw database records, the AI interprets the information and presents it in a format that is meaningful to project stakeholders.

What Makes This Different?

Traditional reporting tools answer questions you thought of when designing the report.

AI-powered reporting answers questions you think of today.

Instead of building dozens of dashboards and reports, users can simply ask:

  • What projects are over budget?
  • Which projects are at risk?
  • What work was completed in the last two weeks?
  • Which projects are overdue?
  • What should I discuss in my project review meeting?
  • Give me an executive summary for my manager.
  • Create a customer-friendly project update.
 
The AI dynamically analyses the underlying SQL data and generates the appropriate response.

A Real Example

One of the first prompts we tested was:

"Help me provide a project update for my manager. Include recent activity, progress and risks."

The agent analysed 20 active projects and automatically generated:

Portfolio Summary

  • Number of active projects
  • Budget utilisation
  • Projects over budget
  • Projects past due
  • Current project health status

Recent Activity

The agent identified completed work across multiple customers including:

  • Server upgrades
  • Tenant migrations
  • Mail migrations
  • Password management deployments
  • Managed services onboarding activities

Project Risks

The system highlighted:

  • Budget overruns
  • Late projects
  • Projects requiring closure
  • Resource consumption concerns
  • Timeline risks

All without a project manager manually reviewing dozens of tickets and project records.

The Time Savings

For most project teams, preparing stakeholder updates involves:

Traditional Approach
Activity
Typical Time
Review project records
20 mins
Check ticket activity
15 mins
Analyse budget utilisation
15 mins
Identify project risks
15 mins
Draft stakeholder update
20 mins
Total
85 mins
AI-Assisted Approach
  • Ask a question
  • Review the generated response
  • Make minor adjustments
  • Send to stakeholders

Typical effort: less than 5 minutes

This represents a significant reduction in administrative overhead while improving reporting consistency.

Benefits for Project Managers

Faster Status Reporting

Generate weekly, fortnightly or monthly project updates almost instantly.

Better Risk Visibility

The AI automatically identifies:
  • Over-budget projects
  • Delayed projects
  • Open project reviews
  • Resource concerns
before they become major issues.

Consistent Communication

Every stakeholder update follows a structured and professional format.

Less Time in Reporting

Project managers can focus on delivery rather than administration.

More Informed Decision Making

Questions can be asked as they arise rather than waiting for reports to be developed.

Benefits for Project Coordinators

Project Coordinators often spend significant time compiling information for project managers and stakeholders.

The AI agent helps by:

  • Consolidating information across multiple projects
  • Providing instant summaries
  • Highlighting missing information
  • Identifying overdue activities
  • Preparing draft communications
This allows coordinators to become more proactive and spend less time creating manual reports.

Why Azure OpenAI?

Azure OpenAI allowed us to securely bring AI capabilities to business data while maintaining control of the solution architecture.
The key advantage was the ability to combine:
  • Enterprise-grade AI
  • Existing business systems
  • On-premises SQL Server databases
  • Microsoft ecosystem integration
 
Rather than replacing existing tools, Azure OpenAI enhanced the value of the data already stored within them.

What's Next?

The latest version of the agent has already expanded beyond project reporting.

Additional capabilities now include:

  • Email intelligence
  • Chat analysis
  • Meeting transcript retrieval
  • Project audit trails
  • Time entry analysis
  • Automated document generation with alltasksIT branding
As the platform evolves, users will increasingly be able to interact with organisational knowledge through natural conversation rather than traditional reporting interfaces.

Final Thoughts

AI is often discussed in terms of future possibilities.

For project management teams, the opportunity exists today. By connecting Azure OpenAI to on-premises SQL Server data, we transformed project information from something that required manual interpretation into something that can be queried conversationally.

The result is:

  • Faster reporting
  • Better stakeholder communication
  • Improved project visibility
  • Reduced administrative effort
  • More time spent delivering successful projects

At alltasksIT, we're continuing to explore practical AI solutions that deliver measurable operational improvements for our clients and for our own teams.

The future of project reporting isn't another dashboard.

It's simply asking a question and getting the answer.

Could AI Save Your Team Hours of Project Reporting?

Imagine asking a simple question and instantly receiving a clear project update covering progress, recent activity, budget utilisation, overdue work, and potential risks.
alltasksIT helps organisations connect AI with existing business systems, including Microsoft technologies, SQL Server, and ConnectWise, to transform operational data into actionable insights.
Complete the form to speak with our team about building practical AI solutions around your existing systems and data.

Author

John Koziaris

CEO/Founder

Founder and Principal of alltasksIT with 30+ years IT experience, John has a broad and varied experience across cloud computing strategies.

John has been successfully realising IT and networking solutions for small to medium businesses for over 25 years.