I work on AI voice agents and internal software used by car dealerships.
Most of what I build sits behind the scenes and deals with real workflows, real calls, and real edge cases.
Also working on my first open source project for a prompt management tool.

My main project.

• AI voice agents currently used by car dealerships
• Handles inbound calls and captures structured service information
• Building a Service Ops platform to improve service workflows and reporting

This is where most of my time goes.

The platform behind Uobo.

• Call ingestion and processing
• AI evaluations and scoring
• Multi-tenant data separation
• Analytics and reporting layers

It handles the infrastructure so products like Uobo can stay focused on the user.

An open-source prompt management tool I started as an internal project.

• Versioned prompts
• Forking and reuse
• Variables for automation workflows
• Private or public libraries

Built for people who actually write and reuse prompts.

• AI evaluations and scoring
• System design for multi-tenant products
• Kubernetes fundamentals
• Building and testing AI systems with low compute

• Systems that work at odd hours without babysitting
• Clear data boundaries
• Readable, boring code
• Ship first, then improve

If something can’t be explained simply, it probably isn’t ready.