AI developer · Application engineer · BI & data
I build AI-connected applications and data systems that make complex work easier to run and understand.
I design MCP integrations that give AI assistants live application context and governed ways to act. In PrototypeX, I also use AI-assisted workflows to build, simulate, diagnose, and refine virtual systems.
01 / Selected work
Three applications addressing complex work in virtual prototyping, system design, and accounting.
Featured project / 01
An interactive environment for exploring physical system ideas. Build a virtual model, configure behavior, run selected tests, inspect results, and revise the design. AI assistance can accelerate the workflow while a person guides the work.
Current status Fully developed and mature. Extensive additional testing is underway, with smaller features under consideration. Version 2 is in planning.
PrototypeX demonstrations
An LLM built these configured models under user guidance in PrototypeX. Each recording shows the resulting simulation running in the application.
A virtual ring spins to create approximately 1 g of apparent gravity at its outer deck. The recording compares attached and free-floating objects with live acceleration readings.
A virtual rocket combines a main motor, control thrusters, sensors, and landing gear in a scripted climb and descent with live telemetry.
A virtual quadcopter uses controlled rotors and onboard sensors for a scripted takeoff, hover, and guided descent.
AI-assisted case study
An AI client used PrototypeX to create an autonomous platform, test a three-waypoint route, diagnose a control limitation, revise the model, and run it again. The useful result is the visible loop from intent to measured evidence.
These are results from a configured virtual model. They do not establish physical performance or engineering certification.
Recorded PrototypeX workflow; playback is accelerated.
02 / System design
A contract-driven platform that turns one accepted relational design into selected data, reporting, application, and documentation deliverables. Its MCP integration lets an AI assistant help refine the live design.
Current status Mature application undergoing additional testing. The deliverables section is planned for future expansion.
03 / Accounting
Configurable accounting software with MCP integration for AI-assisted query and audit, authorized operations, and governed configuration.
Current status Major development work is complete. Further refinement and testing are underway.
System Synthesis Studio workflow
The user describes the system once. Guided questions and an AI design assistant help turn that description into an accepted contract. The user then selects what to generate from that shared design.
Start with a name and short description, then work through several design sections, usually with four to six questions per section. The assistant can help propose and refine answers alongside the live contract.
After accepting the contract, choose a build profile or individual deliverables. The selection can include data, reporting, documentation, databases, and API outputs.
Before building, inspect the selected outputs, dependencies, warnings, and the exact folders where the deliverables will be written.
The results screen groups each generated item and explains how to open it. In the illustrated build, the package includes Power BI assets and sample data; the report has multiple KPI and reporting pages.
These applications are owned by SPACE INTERTECHNOLOGIES LLC.
02 / AI development
I build application integrations that give AI assistants useful context and clearly defined ways to act. MCP connects the assistant to the application; people retain control over changes that need review or approval.
MCP gives an assistant access to the active design so it can help examine requirements, suggest improvements, and work with the same design the desktop app uses.
Its MCP interface separates read-only insight, authorized operations, and governed configuration into distinct modes.
I guide AI-assisted model creation, simulation, diagnosis, and revision in a full application, then inspect the recorded results and the model's limits.
03 / Experience
Selected outcomes from production application, reporting, and data work. Company and client details are omitted here.
Redesigned a long-running reporting and premium calculation process, reducing an overnight run from more than five hours to approximately 90 minutes.
Built a reusable SQL Server migration framework and used it to move thousands of databases while preserving naming and operational continuity.
Built partner feedback ingestion with C# and T-SQL staging, Type 2 history, integrity monitoring, and automated exception notices.
Developed large SSAS Tabular models with automated processing, detailed drill-through, and quick exception analysis.
04 / Approach
I’m drawn to software that helps people model a problem, see what is happening, and make an informed next decision.
Across these projects, that means turning detailed rules and workflows into clear interfaces, preserving evidence of what happened, and keeping people in control of AI-assisted changes.