July 28, 2025
When you walk into Techtree Partners, a 20-person network cabling and security firm in Springfield, Missouri, you might not expect a full-scale data warehouse operation humming beneath the surface. But that’s exactly what Caleb Royer built, without a traditional background in data engineering and without a team.
“I wear several hats,” Caleb says, smiling. “I’m the developer, the data engineer, the guy that manages our ERP system... all of it.”
From Spreadsheet Chaos to Data Discipline
Techtree’s initial approach to data was piecemeal. Power BI reports were stitched together from BigQuery, which itself was populated by replicated data from Odoo, their ERP of choice. Modeling was done directly inside Power BI. “It fell apart pretty quickly once things got complicated,” Caleb admits.
Attempting to evolve, Caleb turned to Microsoft Fabric, hoping their low-code offering could bridge the gap. But once SCD Type 2 dimensions and other warehouse logic came into play, Fabric proved too limited.
“I was going to cobble together a full stack with Spark, DBT, Airbyte, just to let AI control the system,” Caleb recalls. “Then ChatGPT suggested Mage. I clicked the link, and I haven’t looked back.”
The All-in-One Stack That Let Him Scale
What sold Caleb on Mage was the clean UI, integrated orchestration, and the ability to manage workflows in one place.
“I didn’t want to copy and paste between tools or prompt windows,” he explains. “With Mage, the AI can see everything. It understands my context, builds the pipeline, runs it, even monitors it. I don’t have to babysit it.”
While Caleb was new to data engineering concepts like SCD Type 2, he quickly ramped up by combining YouTube DBT tutorials with the Kimball Data Warehouse Toolkit. “I use DBT snapshots for most of the logic, and everything runs in Mage.”
A Real Data Warehouse, Finally
With Mage, Techtree now has a proper medallion architecture pipeline that starts in Odoo, flows through BigQuery via Google DataStream, is modeled in DBT, and visualized with Reflex. The AI helps create and manage Reflex charts as well.
“We’re tracking estimates versus actuals down to the employee level,” Caleb explains. “If we quote 100 hours and go over by 10, that’s a problem. Now we catch it early.”
They also monitor technician log note percentages, log time accuracy, and even use the data to inform future project bids. “It’s all about visibility. We finally have it.”
The Aha Moment
Caleb’s defining moment with Mage came when he needed QuickBooks data, an integration that wasn’t natively supported yet. “I just told the AI what I needed, and it built a custom block. I ran it. Got the data. That fast. I knew then that this was going to work.”
One Person, Full Stack
Techtree Partners doesn’t have a data department. They have Caleb, and Mage.
“I never touched data before two and a half years ago,” Caleb says. “Now I’m running a full warehouse. I can go from a new field in Odoo to a live report in an hour. Before, that was unthinkable.”
He credits the modularity and simplicity of Mage’s orchestration platform for enabling him to move quickly. “I don’t need ten tools or a team. Just Mage.”
Looking Ahead
Caleb’s roadmap includes retiring tools like n8n and Make.com and moving even more automation into Mage. His biggest ask is more prebuilt integrations.
“If Mage had nodes for everything like n8n does, I’d do it all here. No switching context, no bouncing between tools. Just data in, insight out.”
Final Word
Asked to summarize Mage’s impact in one sentence, Caleb doesn’t hesitate:
“Mage lets me see everything, control everything, and manage complex systems without needing to live in the weeds.”
For Caleb and Techtree, Mage didn’t just improve the data workflow. It made it possible in the first place.
