StackSwap is a GTM optimization system
A layer that turns stack signal into interpretation, then into actionable next steps. Below: what it runs on, what’s live, and how the data layer is built.
Under the hood
Tech stack used
Languages, runtimes, and platforms holding everything together—chosen for velocity without giving up safety or cost control.
⚡
Platform.js
Shared nav + footer, metrics, Supabase, count-up animations. Pufferfish.
Read →🔒
Edge API Proxy
Vercel serverless. CORS handling. Server-side API keys.
Read →🎨
main.css Design System
Single CSS source of truth. Cards, buttons, inputs, metrics, pills, grids. All 14 pages.
Read →🧪
Recommendation Normalization Layer
Deterministic validation pass that resolves conflicting KEEP/REPLACE/REMOVE/ADD outcomes before render.
Read →▲
Next.js Hybrid Router
Pages + App Router in one codebase. Marketing on Pages, StackScan results on App.
Read →🧠
TypeScript End-to-End
Strict TS across UI, API, and server code. Shared types for reports, leads, intake.
Read →💳
Stripe Payment Rail
Checkout sessions, webhooks, fulfillment state, idempotency — report-bound, no accounts.
Read →🤖
LLM Multi-Provider
OpenAI + Claude behind a common surface. Structured outputs, token-usage logged per report.
Read →🚀
Vercel Edge + Serverless
Preview deploys per PR. Edge + Node runtimes. Cron + background jobs for StackScan work.
Read →In production
Features shipped
User-facing capabilities and product surfaces that exist today. Each card links to an article covering what it does, how it fits, and why it matters.
🎯
Tool Capability Map
16 tools with functions, strengths, limitations, replacement rules, fit logic.
Read →🚀
Recommendation Engine
7 stack blueprints, 6 fit rules, company-specific stack recommendations via Claude.
Read →📈
AI-Native Gap Metric
Measures legacy exposure as a percentage. Color-coded. Benchmarked against industry.
Read →🔗
Cross-Tool Data Flow
StackScan, StackBuilder, StackEnrich, and GTM Strategy pass context via URL params + sessionStorage.
Read →🧩
StackScan Flow
URL-first intake, Supabase-backed lead capture, and unified GTM Engine results route.
Read →🌐
Company URL Inference
Server-side company lookup API that infers name, industry, positioning, and competitors from a single URL.
Read →📄
GTM Report Template
10-page dark-mode GTM strategy HTML template and TypeScript generator wired for PDF export.
Read →🧱
StackScan Data Adapter
Adapter that maps StackScan / StackBuild analysis and form context into a single GTM strategy report model.
Read →💳
Report Checkout + Unlock Flow
Stripe Checkout for one-time $99 report unlocks, webhook verification, and report-level premium state.
Read →🧬
Operating Profile Inference
Unified profile object for demo + real users, powering tailored executive summary and benchmark credibility lines.
Read →📋
Sortable Plan Table
Tri-state column sort with affordance arrows. Shared component between paid, free preview, and pre-paywall surfaces.
Read →📊
Premium Comparison Pages
13 /compare/[slug] pages with Article + FAQPage JSON-LD, structured verdicts, and editorial tuned for LLM-SEO.
Read →📚
Knowledge Base Engine
Comparison hub, category pages, 60+ articles. FAQ schema + canonical URLs drive LLM-SEO across the entire KB.
Read →Architecture write-ups
Infrastructure & data
Long-form notes for operators and integrators—schemas, seeding, dashboards, shared platform code.
🗂
Supabase Data Layer
21-column schema. RLS policies. Source tagging. Organic intelligence views.
Read →📊
Internal BI Dashboard
Real vs seed filtering. Funnel analysis. Password-gated.
Read →💰
1,120-Tool Database
Per-seat + flat-rate pricing. Team multipliers. Industry benchmarks. 837 priced.
Read →📬
Leads Schema v2
Extended Supabase leads table for StackScan: GTM fields, analysis scores, spend estimates, source tagging.
Read →🎲
StackScan Fuzz Harness
500-scan invariant sweeps across 12 rules. Catches pricing drift, circular logic, and phantom savings before users do.
Read →💵
Canonical Money Architecture
One-number-across-surfaces rule: hero, plan table, stack transformation, and CTA all read the same canonical source.
Read →📏
Team-Size Precision Pipeline
Precise numeric headcount threads through scanStack end-to-end. No more lossy band round-trip between intake and scan.
Read →🔄
Circular Redundancy Safeguards
Dedup logic blocks "remove A because B / remove B because A" rationales from ever reaching the execution timeline.
Read →⚖️
11-Source Weighted Pricing Council
11 independent SaaS pricing datasets merged with confidence weighting behind every tool cost estimate.
Read →