Systems we've designed, built, and shipped

A cloud-based 3D design and catalog platform that lets designers across the group's playground-equipment brands lay out equipment, price it, and push approved designs through to manufacturing — backed by an Azure SQL and Cosmos DB data layer, with a React front end and a .NET API behind it. Architecture and modernization work is ongoing.

• An Azure Data Lake Storage pipeline — Event Grid triggering Logic Apps workflows — ingests designer-submitted DXF, GLB, and PNG files and converts each upload into 2D, 3D, and thumbnail assets automatically
• Postman- and Data-Factory-triggered data operations replaced with a UI-driven application for SQL-to-Cosmos loads, cross-environment SQL sync, and a dynamic SQL Table Manager — removing the need for manual scripts to move data between Dev, Test, and Prod
• A Cosmos DB container model feeds the live design tool: catalog, components, compliance, pricing, and layout data each resolve from SQL through dedicated stored procedures into their own container, keeping reads at single-digit-millisecond latency
• An AI roadmap scoped on Azure OpenAI Service and Azure AI Services — a generative design wizard, natural-language design input, automated compliance validation, and a help-desk assistant — to cut the time designers spend on repetitive layout work

A dedicated data migration application for the group's cloud design platform, replacing three separate manual processes — Postman-triggered Azure Functions, hand-run Azure Data Factory pipelines, and direct database queries — with a single UI the team can operate without touching Postman, the Azure Portal, or SQL directly.

• A UI-driven SQL-to-Cosmos transfer tool: selecting an environment and target container auto-populates the stored procedure, partition key, and request-unit (RU) throughput settings, replacing a Postman flow that required every parameter to be set by hand
• Manual triggering of 95 table-level Data Factory pipelines replaced with a UI that lets a user pick one table — or bulk-select many — and sync Dev→Test or Test→Prod without opening the Azure Portal
• A dynamic SQL Table Manager with full CRUD on any table via an auto-discovered schema, including foreign-key-aware dropdowns that resolve IDs to readable names
• Group-based security (Administration vs. Brand Manager roles) so pipeline and environment access lines up with who's actually supposed to touch production data

A cloud-based order management system built on an Azure microservices layer, integrating external marketplaces such as Shopify through a queue-based accelerator — delivered across three coordinated development pods spanning internal and external teams.

• Azure microservices architecture for e-commerce order management
• Marketplace integration (Shopify and similar platforms) via a queue-based accelerator
• Front-end and back-end delivery coordinated across three development pods

A ground-up Azure platform for a real-time 3D design and configuration application, including the compliance engine that validates every change a user makes as they build.

• Cosmos DB loaded through rules-driven, event-triggered pipelines
• Real-time compliance engine powered by Azure SignalR Service, pushing validation results to every connected client instantly instead of polling for updates
• Serverless Azure Functions and web workers supporting concurrent 3D rendering
• Data collection foundation laid for future generative-design and AI work

The integration layer connecting Salesforce and a nonprofit workflow platform, automating processes that previously required manual handoffs between systems.

• Azure Data Factory pipelines syncing data to Azure SQL Database on a scheduled, trigger-driven cadence
• API integration automating nonprofit workflow, task, and time-entry data
• Salesforce integration across the data platform

An Azure rebuild spanning two of the country's largest casual-dining restaurant groups, replacing 300+ legacy SSIS and SSRS packages with a modern Azure Data Factory and Azure Databricks architecture, delivered by an offshore team of 6 data developers and 2 MVC developers.

• 300+ on-prem SSIS/SSRS packages rebuilt using Azure Data Factory and Azure Databricks' Spark-based ETL engine
• Azure Data Lake Storage designed with hot, cool, and archive tiers to match storage cost to how often each dataset is actually accessed
• Delivered by an offshore team of 6 data developers and 2 MVC developers
• Technical specifications delivered for every migrated solution

A hub-and-spoke on-prem data model redesigned into an event-driven Azure platform, connecting Microsoft Dynamics CRM, field services, Salesforce, and Yardi into a single source of truth for the business.

• Event-driven architecture using Logic Apps, Azure Functions, and Event Hubs to move data the moment it changes instead of on a fixed schedule
• Azure SQL Database hub supporting ETL, integrations, and BI reporting
• Design, data, and backend API development coordinated across dedicated teams

Two engagements, each recovering a stalled delivery timeline: automated federal and state regulatory data calls for one client, and a policy-and-claims platform for the other that went from live in 4 states to 39 in 60 days.

• Automated regulatory data calls across multiple states (TX, NV, AR, AZ)
• Legacy FoxPro/SQL Server systems migrated into Azure Data Factory pipelines
• Policy & claims platform delivered in 60 days, live in 39 states

Automated a weekly and monthly "data tape" process that had been entirely manual — CSV exports from the property-management platform loaded into a macro-driven Excel workbook to hand-build the investor-facing KPI reports. Replaced it with an Azure Data Factory pipeline suite pulling general ledger, rent roll, delinquency, tenant, and property data directly from the source system's API into Azure SQL Server.

• A config-driven master pipeline loads all active data feeds nightly in priority order — general ledger, rent roll, delinquency, tenant and property directories, unit turns, tenant tickler — each landing through a merge stored procedure that inserts or updates rather than reloading everything
• A dedicated month-end close pipeline stages cash and accrual general ledger files through Azure Data Lake before merging into SQL Server, with the run promoted from Dev through QA to Production
• Automated the investor-grade reporting package institutional owners require on a fixed cadence — occupancy, leasing and traffic, NOI by property, delinquency, and trending KPI reports — replacing hand-built Excel workbooks
• New portfolios and data feeds onboard by adding a row to a config table rather than rebuilding a pipeline from scratch

A real-time 3D car configurator that lets racing teams preview sponsor branding before it goes on the car — paint schemes, decals, and logo placement across the exterior, interior, and driver's uniform, rendered live and viewable from a track-side camera angle or the driver's-eye view.

• Three.js/React Three Fiber rendering with full 360° rotation, swappable manufacturer body models, and a sponsor-logo library that applies to exterior, interior, and uniform positions in real time
• A design-submission and approval workflow: a team builds one or more designs, submits them as a package, and the sanctioning body reviews and approves or rejects each with comments, all over email notification
• Cloud-agnostic architecture — a .NET Core API and blob storage for 3D assets sit behind a framework that can move between Azure, AWS, or GCP without a rewrite
• Delivered a working demo covering every track on the series calendar, team-to-manufacturer model mapping, and role-based access (Admin, Designer, Viewer) for stakeholder review

An exploratory build that shipped all the way to the app stores — a secure, cross-platform messaging app published on the Apple App Store and Google Play as an alternative to the stock texting app, built end-to-end on a single Azure-native stack.

• One codebase across iOS, Android, Windows, and macOS via .NET MAUI Blazor Hybrid, instead of maintaining separate native apps per platform
• Real-time messaging over a SignalR hub backed by Azure Cosmos DB, giving one-on-one and group chat, typing indicators, and read receipts sub-second delivery at global scale
• Voice and video calling built on Azure Communication Services, with call history and missed-call alerts alongside the messaging experience
• Push notifications unified through Azure Notification Hub, dispatching to FCM and APNs from a single backend call regardless of platform
• Media and file sharing through Azure Blob Storage and CDN, with time-limited SAS URLs so shared files are never exposed on a public endpoint

A ground-up rebuild of what a phone's default texting app should be: the people you actually talk to float to the top as bubbles, business texts sort themselves into topic-based lanes, and unknown numbers wait in a Screener until you say yes — all running entirely on-device, with no servers, accounts, or ads.

• Built on .NET MAUI targeting Android, sharing one codebase and design language with AzureForge's other cross-platform work
• Three switchable home-screen layouts (bubbles, bubbles + color list, or color list) with user-defined color filters for family, teams, and work
• Smart Chats/Updates/Screener lanes classify incoming SMS/MMS automatically — verification codes still notify with one-tap copy, everything else sorts silently
• Zero backend: every message, contact, and preference lives in the phone's own storage — nothing is transmitted anywhere except the text itself, over the carrier, like any SMS

See ForgeLink SMS & buy

A tool that turns any public storefront on a popular handmade-and-vintage marketplace into a source for instant marketing: browse a seller's full catalog, pick a listing, and generate a polished standalone landing page — rewritten headline, story copy, and benefit highlights — grounded strictly in that listing's real data, with platform-tailored social post drafts ready to publish.

• One AI call produces the full landing page copy and one social post draft per platform (TikTok, Instagram, Facebook, Pinterest), each written to that platform's own voice, length, and hashtag conventions via schema-validated structured output
• A strict no-fabrication system prompt keeps every claim grounded in the listing's actual title, description, price, materials, and tags — elevating the pitch without inventing details
• Direct publishing integrations for Pinterest, Facebook, Instagram, and TikTok, each handling that platform's own auth model — long-lived tokens for Meta and Pinterest, a full OAuth login flow for TikTok
• Admin-gated accounts and role-based access on top of a real user database, so the tool ships as a controlled multi-user product rather than a single-user script

A portfolio dashboard that pulls a client's full financial picture into one view — every linked brokerage, retirement, and cash account, rolled up into a single real-time net-worth figure and broken out by account type, with historical value trends per category and per holding.

• Securely links brokerage and retirement accounts read-only through Plaid, with encrypted access-token storage and a reconnect flow when a linked account's token expires
• Every holding is auto-categorized into Retirement, Taxable Brokerage, or Cash & Other based on the account's own subtype — no manual tagging required
• A daily scheduled job snapshots every linked account across every client, building the historical time series that powers per-category and per-symbol value charts over time
• A sandbox/production environment toggle lets the team test against Plaid's sandbox data without ever touching a real linked account