I build production-minded software for complicated business workflows—AI agents, decision systems, operations platforms, pricing engines, APIs, and SaaS applications.
My strongest work sits where software has to do more than look good: it has to enforce rules, explain decisions, recover from failures, protect data, and remain maintainable after the first release.
- Core engineering: C#, .NET, Python, Node.js, TypeScript, React, Next.js
- AI systems: LLM applications, agents, RAG, tool calling, structured outputs, LangGraph/LangChain
- Cloud and data: Azure, SQL Server, PostgreSQL, Docker, CI/CD
- Architecture: distributed workflows, queues, idempotency, observability, rule engines, multi-tenant SaaS
- Based in: Miami, Florida
- Open to: Senior, Lead, Staff, and AI/full-stack engineering opportunities
| If you want to see… | Open this |
|---|---|
| A deterministic business rules engine with live proof | Supplier Pricing Engine · Live demo |
| A resilient full-stack logistics architecture | Logistics Control Tower |
| A working product with formula-level audit checks | DealCheck Pro |
| How I structure production-minded portfolio work | Builder's Desk Labs |
A deterministic Configure–Price–Quote architecture for products governed by different supplier rulebooks.
- versioned supplier rules and price grids
- dimension rounding and compatibility validation
- landed cost, margin, discount, approval, deposit, and balance calculations
- explainable calculation traces and quote provenance
- automated regression tests and CI
- interactive multi-room quote workflow with PDF output
A full-stack logistics operations demonstration covering last-mile optimization, delivery-state integrity, capacity-aware load building, failure recovery, and live operational events.
- React/TypeScript operations interface
- Node/Express API and Server-Sent Events
- Python/FastAPI optimization service
- TypeScript fallback optimizer for graceful degradation
- deterministic incident simulation
- automated domain tests, CI, and Docker Compose
A portfolio of focused engineering case studies designed around real operational and commercial problems rather than generic tutorial applications.
A modern Next.js application with a documented execution plan, typed frontend architecture, deployment configuration, and product-focused implementation.
A working Excel-based residential real-estate screening model.
- rental and flip strategy comparison
- configurable investment thresholds
- formula-driven offer ceiling
- 25-cell rental downside matrix
- four flip downside scenarios
- six visible model-integrity checks
- downloadable working workbook with documented limitations
My public GitHub history dates back to 2018. The repositories below are retained intentionally to show how my work progressed—not to present old dependencies as current production choices.
| Period | Engineering focus | Representative evidence |
|---|---|---|
| 2019 | C#, .NET services, Xamarin, Android/iOS, Azure data access | Existing_DotNet, Cryptoquick, CQ-Orderbook, CryptoQuickApp |
| Later platform work | APIs, automation, financial applications, integrations, and full-stack product experiments | Selected original repositories remain public as historical engineering evidence |
| Current | AI workflows, resilient distributed systems, deterministic business engines, logistics, SaaS, and production architecture | Supplier Pricing Engine, Logistics Control Tower, DealCheck Pro |
The historical projects now explain what they demonstrate, their support status, and how I would modernize them today.
This account includes older experiments, archived-era mobile and crypto work, tutorials, and upstream repositories retained as part of my development history.
Unless a repository explicitly identifies my contribution, do not treat a fork, sample, clone, or tutorial as original authorship. My current original portfolio work is presented in the Featured engineering work section above and documents what is implemented, what is simulated, and what I personally designed.
I start by mapping the existing system: business rules, data flow, failure modes, bottlenecks, dependencies, and production risks. A rewrite is a business decision—not a reflex.
AI can extract, classify, recommend, and orchestrate. It should not silently become the authority for a price, permission, payment, compliance decision, or other result that must be reproducible and auditable.
Production workflows need explicit state transitions, unique operation identifiers, idempotency, retries, dead-letter handling, observability, and safe recovery—not just a successful happy path.
Strong software should explain itself through clear boundaries, typed contracts, tests, decision records, operational telemetry, and documentation that helps the next engineer make safe changes.
I am currently building and refining:
- autonomous and human-in-the-loop AI workflows
- RAG systems with grounded, traceable answers
- enterprise AI services using Python/FastAPI and Node/.NET
- logistics and operations-control systems
- configurable pricing and rules platforms
- secure, observable, production-ready SaaS applications
If your team is taking complex ideas from zero to production, I would be glad to talk.