Kelly Tweets — Pipelines & Workflows

Date Compiled: 2026-04-27
Summary: These tweets document the evolution of Kelly's factory pipelines — from ad-hoc parallel builds to a structured 10-step iOS pipeline (DISCOVER → VALIDATE → DESIGN → BUILD → POLISH → SUBMIT → MARKET → LEARN) and the emergence of specialized sub-factories: Design Station, Marketing Factory, and Research Station. Each factory module became a discrete, repeatable, improvable component of the overall system.

Key Concepts

  • iOS Factory 10-step pipeline: 0-CHECK STATUS, 1-DISCOVER (opportunity-scanner), 2-VALIDATE (28/40 gate), 3-DESIGN (77/110 gate), 4-BUILD, 5-POLISH (70% gate), 6-LAUNCH, 7-SUBMIT, 8-MARKET, 9-TRACK, 10-LEARN
  • Design Station: Crowd-sources design trends from X, generates specs with Gemini, auto-applies styles to any app. One prompt → 10 agents → design on demand.
  • Research Station: Scrapes Reddit + App Store for opportunities, scores niches, identifies red oceans vs. blue oceans. Factory learns from its own competitive analysis.
  • Marketing Factory: End-to-end automated marketing per app — landing pages (real copy), Twitter threads (5-6 tweets), TikTok scripts (3-4 videos), Instagram carousels + captions. First to Influencer Marketing automation.
  • Marketing pipeline gap: "Being able to automate the marketing pipeline is what I'm thinking about 24/7." — Austen noting the marketing factory as the strategic moat.
  • Discovery gate: DISCOVER → scan Reddit + App Store → opportunities. VALIDATE → score 28+/40 required before proceeding.
  • Design gate: 95/110 score required before BUILD. Quality bar applied before and after building.
  • Polish gate: 70% threshold before SUBMIT.
  • Factory v3.0: New modules added — ASO (keywords, screenshots, competitor analysis), Validation (scrape reviews, extract real problems), Polish (haptics, confetti, spring animations).
  • Test infrastructure: 100-point scoring system, automated Maestro UI tests (taps buttons, verifies state, checks persistence), design checklist.
  • Self-improving loop: Every app captures wins/failures, runs post-review retros, feeds learnings to next build.
  • Workflow reorganization: Scripts reorganized into gates/, tools/, utils/ with .factory/ state tracking.

Notable Patterns

Pipeline scripts as quality enforcement: Shell scripts as quality gates — an enforcement layer AI can't edit its way around. "An enforcement layer AI cannot bypass." This structural separation between builder and gatekeeper is a recurring theme.

Learning loop: "Built a learning loop for my app factory. Every app now captures wins/failures, runs post-review retros, feeds learnings to the next build."

Scale ambition: "20 working + selling apps per day" — mapping bottlenecks across design, build reliability, distribution, ASO, testing, and growth loops. Turning it into a concrete operating system overnight.

Parallel building at scale: "All 20 apps building and running now." Later: "24 products in 24 hours — not landing pages, not mockups, real production-ready products with full-stack apps, AI integrations, payment systems, live demos."

The factory-to-distribution insight: "As software gets cheaper, distribution will matter." Kelly pivoted to automated faceless social media channels and full marketing packages alongside every build.

ios-factory-pipeline, design-station, marketing-factory, research-station, discovery-gate, design-gate, polish-gate, quality-gates, learning-loop, automated-testing, factory-v3