Old School to New School: The Evolution of Automation from VBA to AI
By Thomas B. Overton
Introduction
In my previous post, From Clipboard to Cloud, I shared the journey of transforming manual, paper-based workflows into automated systems across environmental engineering and federal operations.
But that journey didn’t happen in an era of cloud platforms, AI, Copilots, or scalable data services.
It happened in the 1990s and early 2000s—when automation wasn’t just difficult, it was fragile, isolated, and hard to scale.
Back then, automation required ingenuity. Today, it requires orchestration.
—
The Reality of Old School Automation
When building systems in the early years, every layer of automation came with constraints:
- No cloud infrastructure or centralized data platforms
- Heavy reliance on desktop applications like Excel, Access, and Word
- Tight coupling between data, logic, and document output
- Manual distribution and storage of generated reports
Technologies like VBA Forms and Scripts and Word Mail Merge became the backbone of innovation during that time, enabling users to automate repetitive tasks directly within Office applications.
But these solutions came with trade-offs:
- Scripts and Forms were difficult to maintain and scale
- Automation relied on local machine connections and user sessions
- Systems were not designed for distributed or enterprise workloads
We made it work—but every manual step forward required overcoming platform limitations.
Mail Merge: The First Wave of Document Automation
One of the most powerful tools of the era was Word Mail Merge.
At its core, mail merge connected structured data (typically Excel or Access) to document templates, enabling batch generation of engineering reports; I built hundreds of those of increasing complexity.
- Data lived in spreadsheets
- Templates contained merge fields
- Logic was minimal and often brittle
- Output required manual handling and distribution
This approach worked well for simple scenarios, but as complexity increased, the system began to break down.
Modern document workflows highlight these same challenges—data variability, conditional logic, and scaling needs quickly push traditional merge approaches beyond their limits.
Still, at the time, combining Excel + Word + VBA was revolutionary. It allowed us to:
- Generate hundreds and hundreds of reports
- Standardize document formatting
- Reduce manual effort dramatically
It was the foundation of everything that followed.
The Breaking Point: Scaling Beyond the Desktop
As systems grew, so did the problems:
- Mail merge processes became increasingly fragile
- Complex documents required workarounds and custom scripts
- Large-scale generation introduced performance and reliability issues
Traditional desktop-based automation was never designed for:
- High-volume document generation
- Cloud or server-based execution
- Multi-user, distributed workloads
Lack of scaling and observability created a gap between what businesses needed and what the technology could reliably deliver. And what an individual could operate and monitor.
New School: Fabric, AI, and Intelligent Automation
Fast forward to today, and the paradigm has completely shifted.
Instead of isolated desktop workflows, we now operate within connected, cloud-native ecosystems.
Platforms like Microsoft Fabric bring together data engineering, analytics, and automation into a unified environment.
At the center of this transformation is M365 Copilot a generative AI assistant embedded directly into the data platform. And the Rayfin SDK launched at Build 2026 earlier this month.
These apps and agentic workflows enhance how users interact with data by enabling natural language input, generating insights, and assisting with automation workflows. And for monitoring and reporting.
From VBA Scripts to Intelligent Systems
What once required thousands of lines of VBA and multiple forms to bind locally can now be achieved through integrated, intelligent systems with a Fabric SQL Database and backend services.
Consider the evolution:
- Then: VBA macros triggered manually within Excel or Word
- Now: Fabric pipelines orchestrate data flows end-to-end
- Then: Mail merge templates with limited logic
- Now: Dynamic document generation in multiple formats driven by real-time data and AI
- Then: Manual data validation and QA processes to find anomalies and errant data
- Now: AI-assisted validation, anomaly detection, and exception handling
This is more than an upgrade—it’s a complete redefinition of automation architecture.
Fabric Workspace Apps + Copilot-Assisted Document Automation
One of the most exciting advancements is the emergence of Fabric Workspace Apps combined with AI-assisted workflows.
These capabilities enable:
- Centralized data and business logic in a Fabric workspace
- Integrated pipelines for data ingestion and transformation
- AI-assisted document generation using natural language prompts
- Scalable, repeatable automation across teams and organizations
Instead of pushing data into Word templates, we now:
- Pull insights directly from governed data platforms
- Generate documents dynamically based on context
- Use AI to assist with content suggestions, formatting, summarization, and narrative creation
Copilot acts as an intelligent layer that augments—not replaces—human expertise, improving productivity while maintaining oversight.
What This Means for Architects and Builders
The shift from old school to new school isn’t just technological; it’s architectural.
Success today requires thinking beyond individual tools and toward systems:
- Design for scalability from day one
- Separate data, logic, and presentation layers
- Leverage AI as an augmentation layer, not a shortcut
- Build platforms – not scripts
The mindset has evolved from “How do I automate this task?” to:
“How do I design a system that automates itself?”
—
Final Thoughts: Full Circle
Looking back, the tools have changed—but the mission hasn’t.
From Excel and VBA macros to Fabric and Copilot, the goal remains the same:
- Reduce manual effort
- Improve data quality
- Scale knowledge and impact
The difference is that what once required years of custom engineering can now be achieved with integrated platforms and intelligent assistants in months. Maybe even weeks…
Dare I suggest days!
Old school automation built the foundation.
New school automation builds the future.
Recovery Note
This article is part of a restored series following a site issue. If you missed Part 1, it details the foundational journey from manual processes to early automation systems.
