Amber DunnEngineering Manager · Product & Platform
← All work
Gaming technologyConcept to production: 2016

Building a Casino Operations Platform from a Blank Page

At NexGaming, I took a custom casino operations platform from paper sketches and evolving stakeholder conversations through architecture, full-stack development, third-party integration, testing, deployment, training, production support, and handoff.

The project began with little more than a client request and a series of conversations. No existing product met the casino’s combined needs, and there was no settled specification, design, architecture, or implementation plan. Over nine months, I turned that ambiguity into a working production platform supporting vouchers, gaming activity, employee workflows, customer purchases, permissions, reporting, transaction history, and operational accountability.

Complexity becoming clarity

From scattered operations to one shared platform

Before

SpreadsheetsEmailManual approvalsSeparate toolsFragmented reportingHidden ownership

After

Shared operations platform
  • Role-based workflows
  • Centralized data
  • Clear ownership
  • Operational visibility

Stakeholder convergence

One platform, multiple operational realities

Operations
platform
Casino operationsFinanceComplianceGaming-system partnerEngineeringLeadership

Ambiguity to delivery

Creating direction from a blank page

  1. Discovery
  2. Requirements synthesis
  3. Architecture
  4. Phased delivery
  5. Operational adoption
Decision principlesUnderstand the operation before designing the interface.Separate shared workflows from role-specific needs.Design for change, not only launch.

The impact path

Inherited

The project began as an idea and a client request rather than a defined product, specification, architecture, or implementation plan.

Owned

I led product discovery and business analysis, translating stakeholder conversations into workflows, permissions, data relationships, and implementation decisions.

Changed

I began with paper sketches, data models, query designs, and flowcharts so the business problem became understandable before it became code.

Result

9 months — Concept to production

What I inherited

The problem and the reality around it

The problem

  • The project began as an idea and a client request rather than a defined product, specification, architecture, or implementation plan.
  • Each stakeholder meeting revealed additional workflows, integrations, exceptions, reporting needs, and operational concerns, so discovery and delivery had to evolve together.
  • The platform needed to connect customer kiosks, employee tablets, voucher hardware, gaming systems, permissions, financial activity, and reporting in one accountable operational system.

The constraints

  • Requirements continued to evolve throughout the build, making modularity and versioned technical plans more practical than pretending the original scope was fixed.
  • I developed and launched the platform remotely without visiting the casino, so observability, guided testing, documentation, and production visibility were essential.
  • Barcode-library conflicts and an external gaming-system integration introduced the most difficult technical problems.
  • Financially sensitive employee actions required role-based access, traceability, transaction history, and clear audit records.
  • The initial release served one casino location; early white-label scaffolding existed, but a complete multi-tenant product remained a future direction.

What I owned

My responsibility in the work

  • I led product discovery and business analysis, translating stakeholder conversations into workflows, permissions, data relationships, and implementation decisions.

  • I designed the system architecture, paper wireframes, data models, queries, flowcharts, frontend structure, and backend services.

  • I built the complete initial product across the MongoDB, Express, AngularJS, and Node.js stack.

  • I owned testing, deployment, remote implementation, staff documentation, onboarding, training, and the first 30 days of live production support.

  • I prepared the codebase and operating knowledge for a direct handoff through code documentation, architecture notes, setup guides, walkthroughs, and training sessions.

How I approached it

Decisions, tradeoffs, and delivery

My recurring approach was to understand the problem, model the system, and then implement the solution. I sketched interfaces on paper, mapped data relationships and queries, and used flowcharts to understand how information and transactions needed to move. Because the requirements could not be frozen, I optimized for modularity, reusable components, focused feature branches, and technical plans that could be revised honestly as the product evolved. Because I could not be on-site, I built observability, super-administrator access, notifications, error handling, documentation, and testing into the delivery strategy rather than treating them as follow-up work.

  1. I began with paper sketches, data models, query designs, and flowcharts so the business problem became understandable before it became code.
  2. I divided the work into focused features and feature sets, maintained a strict folder structure, and emphasized modularity and reuse as requirements changed.
  3. I built the voucher lifecycle from unique barcode generation and kiosk printing through game-floor redemption, balance reduction, cash-out, reissuance of remaining credit, and traceable activity history.
  4. I implemented employee authentication, progressively broader role permissions, customer purchases, staff time management, transaction records, live-data tables, and operational reporting.
  5. When vendor documentation did not match real system behavior, I worked directly with the external developer to isolate the mismatch, resolve the integration, and clarify the documentation for that use case.
  6. I created super-administrator access, notifications, error handling, live production visibility, and a broad testing suite covering backend logic, APIs, frontend workflows, and end-to-end voucher behavior.
  7. I guided casino staff through physical testing and implementation over video, explaining which device to use, what action to perform, and what result to verify.
  8. When the voucher system failed around 3 a.m. during production support, I traced the incident to the barcode library and pushed an immediate fix.

How I led

The team and stakeholder system

Although I was the sole developer for the initial product, the work depended on close collaboration. I worked iteratively with casino stakeholders to discover workflows, guided staff through physical implementation and testing over video, and partnered directly with an external gaming-system developer when documented behavior did not match the real integration. After launch, I transferred the working system through code documentation, architecture notes, setup and onboarding guides, walkthroughs, and training sessions; the next team took ownership without an extended dependency on me.

Results

What changed

9 months

Concept to production

The platform progressed from the first stakeholder meetings and paper models to a working production launch. · My contribution

1 developer

Initial product ownership

I designed and built the complete first production release across discovery, architecture, UX, frontend, backend, testing, deployment, and support. · My contribution

30 days

Live launch support

I supported the platform through its first month in production until calls subsided and the system became part of normal operations. · My contribution

Leadership evidence

How I moved the people system

  • — I created structure inside a continuously evolving problem and kept stakeholders moving toward a usable product rather than waiting for perfect requirements.
  • — I remained resourceful, persistent, and calm through third-party integration problems, remote implementation, and live production incidents.
  • — I translated business language into product decisions, architecture, training, and an operating system that people could use in the real world.
  • — I made the system transferable by documenting the code, architecture, setup, onboarding, and operational knowledge before walking away from the project.

Technical evidence

How I moved the product system

  • — End-to-end MEAN-stack architecture and implementation for the complete initial production product.
  • — Voucher lifecycle, employee authentication, role-based permissions, transaction logging, audit history, reporting, and live operational data.
  • — Remote observability and testing across backend logic, APIs, frontend workflows, hardware-assisted scenarios, and end-to-end voucher behavior.
  • — Resolution of conflicting barcode-library requirements and a vendor integration mismatch between documentation and actual behavior.
  • — MongoDB-backed application deployment through Heroku, with domain and DNS management through GoDaddy.

Technical footprint

Technologies and system areas

MongoDBExpressAngularJS 1.xNode.jsJavaScriptHerokuGoDaddy DNSBarcode generationThird-party integrationsRole-based access controlReporting and audit trailsAutomated testing

What I took from it

The lesson I carry forward

What surprised me most was watching the project evolve from paper into production: scattered sketches, models, and conversations gradually became a real system people depended on. I am proud that it launched, stabilized, and worked. The experience demonstrated that I can move across disciplines, create structure where none exists, stay resourceful and calm under pressure, and remain accountable for the whole outcome. Today, I would pair that adaptability with stronger requirements gathering, major milestones defined at kickoff, explicit scope control for future feature ideas, and additional engineering support. Being capable of doing everything does not mean one person should be expected to do everything.

Related work · Technical discovery