CostRadar.ai — Profitability OS

Background E-commerce teams drown in revenue dashboards while true margin stays hidden in spreadsheets. Platform fees, app subscriptions, shipping, COGS, and paid media live in silos. ROAS looks healthy; net profit does not. CostRadar.ai was built as the missing Profitability OS: one place to see what merchants actually keep after every cost line. Challenge 1) Unify Shopify, Amazon, and WooCommerce into a single trustworthy P&L. 2) Match costs to orders and SKUs automatically — not monthly CSV archaeology. 3) Move decisions from ROAS to POAS without enterprise pricing. 4) Ship a polished, production-grade product as a solo founder against well-funded attribution tools (Triple Whale, BeProfit, etc.). Execution Product and UX were designed around a profit-first workflow: connect stores → sync costs → see net margin → act on anomalies and unprofitable campaigns. Stack: React + TypeScript frontend; Supabase Postgres, Auth, and Deno Edge Functions; OpenAI for AI briefing/insights; native integrations with Shopify, Amazon Seller Central, WooCommerce, and major ad platforms. Development accelerated with Cursor (vibe-coding) while keeping architecture, security, and billing production-hardened. Shipped surfaces include Home Dashboard, Costs & Cost Intelligence, Calendar, Orders profitability, Metrics command palette, Anomalies, Simulations, Audiences (RFM), Inventory Intelligence, Pulse multi-store dashboard, AI Chat/Briefing, REST API + MCP, and Zapier. Results A live commercial product with free trial, flat mid-market pricing, public demo, and full tri-platform parity for sync/status/inventory reads. Differentiation roadmap (POAS, savings loops, portfolio P&L, health score, inventory profitability, DTC benchmarks) is implemented on production backends. The work proves a complex B2B fintech-style dashboard can be designed, engineered, and launched by a solo builder using modern AI-assisted workflows — without sacrificing UX clarity or operational depth. Lessons Profit visibility beats vanity metrics. Merchants and agencies need one P&L across channels more than another attribution graph. AI helps ship faster; product judgment still decides what ships.

Shipped a production SaaS at costradar.ai: • Tri-platform sync: Shopify + Amazon (23 marketplaces) + WooCommerce in one account • Full cost stack auto-matched per order/SKU: fees, apps, shipping, COGS, Meta/Google spend • Profit-first analytics: live P&L, POAS (profit on ad spend), anomalies, simulations, inventory intelligence • 54 dashboard widgets, 65+ KPIs, interactive reports, budgets, MCP tools for Cursor/Claude • Public interactive demo: costradar.ai/demo • Pricing from Free → Advanced ($29.99) → Premium ($49.99) with AI Insights Built end-to-end as a solo founder with AI-assisted engineering (Cursor) — product, design, backend, and GTM assets ready for agency pilots.

DTC brands, Amazon sellers, and multi-store operators on Shopify, Amazon, and WooCommerce — plus finance teams and performance agencies that manage those merchants. They need true net profit after platform fees, apps, shipping, COGS, and ad spend. Today most dashboards optimize ROAS and revenue while margins quietly compress. CostRadar turns fragmented store and ads data into one profit layer they can act on daily.

Shipped a production Profitability OS solo: tri-platform P&L, 65+ KPIs, AI insights, MCP for Cursor/Claude — solving margin blindness where attribution tools stop.