From Zero to Production-Grade Platform
ActualPlay.World started the way the best products do: with a very specific gap that became impossible to ignore.
There are thousands of groups of tabletop role-playing game fans making their own podcasts & video shows—known as “actual plays”—but discovery is fragmented, inconsistent, and often biased toward a handful of dominant names. There was no structured, discovery-first platform that treated the subject properly—something like IMDb for RPG actual plays. That insight drove the initial concept and every decision that followed: data modeling, UX, ingestion pipelines, and monetization strategy.
Product Thinking, Not Just Execution
APW was designed as a system, not simply a website.
The core challenge wasn’t just “list shows,” it was:
- How do you represent Shows vs Channels in a way that maps to real-world ambiguity (podcasts vs campaigns vs networks)?
- How do you create a review system that avoids review-bombing while still being useful?
- How do you support ownership and verification of content creators?
- How do you ingest and normalize data from multiple external ecosystems?
Those questions led to a deliberately structured data model and UX:
- Shows and Channels as distinct but flexible entities
- Production-quality ratings instead of generic star systems
- A submission + moderation pipeline designed for scale
- A future-ready ownership/claim system
This level of upfront systems thinking is where most products either succeed or quietly fail.
Design: Thoughtfully Iterated
Every interface in APW was designed from scratch and iterated extensively in Figma, entailing a custom design system emphasizing image-heavy but highly readable layouts.
The design process leaned heavily on rapid iteration:
- Figma as the single source of truth
- Tight loop between design and implementation
- Continuous refinement based on real data and user feedback
Development: AI-Augmented, Not AI-Replaced
APW was built quickly—but not carelessly.
The stack:
- Next.js (via MakerKit) for a production-ready foundation
- Supabase for database, auth, and backend primitives
- Vercel for rapid deployment
- AWS S3 + CloudFront for asset delivery
- Cursor + Figma MCP for accelerated implementation
- Resend for transactional & marketing email
The key distinction is how AI tools were used.
Cursor and Figma MCP weren’t used to “generate the app.” They were used to:
- Translate design intent into code faster
- Reduce boilerplate and repetition
- Keep iteration velocity high without sacrificing control
All architecture decisions, system design, and integration logic were explicitly authored and directed.
Infrastructure: Built Like It Matters
Even at an early stage, APW was built with production-grade infrastructure in mind.
Asset Delivery and Performance
Images are stored in S3 and delivered through CloudFront, enabling:
- Global CDN distribution
- Low-latency asset delivery
- Fine-grained caching control
- Custom image optimization pipelines
DNS and Routing
Rather than relying on basic registrar DNS, APW uses a more robust setup:
- Dedicated DNS management via CloudFront
- Clean separation between application, assets, and email infrastructure
Proxying and Control
CloudFront also serves as a strategic proxy layer:
- Shields origin infrastructure
- Enables rate-limiting and bot mitigation
- Provides a path to advanced edge logic where needed
Data Ingestion: A Real System, Not a Script
A major part of APW’s value comes from its ingestion pipeline.
Rather than one-off scraping scripts, the system is designed to:
- Support multiple ingestors (YouTube, podcast platforms, etc.)
- Normalize incoming data into a consistent schema
- Deduplicate intelligently
- Track provenance of all data
- Allow human review before publishing
Why InventBuild.Studio Is the Right Partner
APW is a representative example of how InventBuild.Studio operates:
- End-to-end ownership from concept to deployment and maintenance
- Deep product thinking, not just implementation
- Fully custom design with rapid iteration cycles
- Modern, AI-augmented development workflows
- Production-grade infrastructure from day one
- Systems designed for scale, not just launch
This is not about building apps quickly. It’s about building the right system quickly—and making sure it holds up when it matters.