I'm a full-stack and AI software engineer with an independent quantitative research practice. These case studies distinguish live products, inspectable public code, and private research implementations—so you can see what is evidenced and what remains confidential.
No imaginary client metrics, false production benchmarks or proprietary signal disclosures.
Select a discipline to explore. Every case includes an evidence link or a disclosure about why its source cannot be public.
Property marketplace & listing portal
A property-discovery experience connecting listings, agencies and prospective buyers. The public site exposes property categories, location browsing, agent pages and posting entry points. My CV describes work across frontend, backend, database, API integration, debugging and deployment.
Commerce & cooperative services
A customer-facing business website combining a shopping catalogue, cart/checkout entry points, membership-oriented services and content. My CV lists KODEKO e-commerce and ordering among delivered web projects.
Laravel application with operational workflows
A Laravel 11 project whose README describes photo attendance, geolocation, leave/permission requests and admin attendance review. The repository exposes the application structure, routes, migrations and tests.
Next.js application backed by Prisma
An educational interest-and-aptitude application with a visible Next.js 15 / React 19 codebase, Prisma schema, middleware and application routes. Demonstrates modern full-stack application structure.
A local-first computer vision service
A FastAPI application that validates uploaded image content, handles EXIF orientation, runs rembg/ONNX inference and returns a ZIP of transparent PNGs. Includes a browser interface, CLI folder workflow and Dockerfile.
Automated media workflow & quality gates
A Python-based content pipeline with asset preparation, transcription, rendering controls and staged validation. Repository artifacts document checksum verification, smoke checks and rollback-oriented patching.
Deterministic research workflow
A local FastAPI/Jinja research MVP integrating historical data imports, a constrained strategy DSL, a replayable backtest engine, ledger/equity artifacts, and evidence-oriented result views.
Cross-exchange data and paper feasibility
Research software for examining funding, fees, borrow availability, liquidity and the constraints of hedged positions. Its documented monitoring mode uses public market data and a separate simulated ledger.
Historical simulation with explicit limitations
A private SOLUSDC 15-minute research strategy with a public monthly/yearly aggregate report, directional statistics, and disclosures about unresolved reconciliation. The algorithm and individual fills are not disclosed.
No case studies match this filter.
Requirements, constraints, security boundaries, and acceptance criteria.
Readable software, appropriate architecture, and testable interfaces.
Tests, reproducible artifacts, operational documentation, and candid limitations.
Custom full-stack systems, AI integrations, research automation and backtesting infrastructure. Share your brief; I'll help scope an achievable first milestone.