DisputeIQ.

DisputeIQ is an AI-first chargeback revenue protection platform for SMB to mid-market e-commerce merchants — Shopify Plus stores ($500K–$10M GMV), Amazon FBA sellers ($2–$50M sales), subscription box companies, and digital goods sellers — automating the entire chargeback lifecycle from pre-ship risk scoring through processor-ready evidence package submission. The wedge against Chargeflow (the closest competitor): $99–$599/month vs Chargeflow's $499+ minimum, AI-first fully automated vs Chargeflow's AI + human review hybrid, SMB / mid-market focus vs Chargeflow's enterprise focus.

Four core features cover the full chargeback lifecycle: (1) Predictive AI Before Orders Ship — flags high-risk orders pre-fulfillment using fraud signals (device risk score, IP/country match, new customer flag, order velocity, AVS/CVV match); (2) Auto Evidence Compilation — aggregates evidence from Shopify, Stripe, UPS/FedEx tracking, customer service tickets automatically per the reason code; (3) Win Probability Prediction — scored per dispute (SUBMIT / ABANDON / REVIEW / PENDING) so the merchant only fights what's winnable; (4) Generated Evidence Package — processor-ready submission in the format the card networks expect.

The persona is "Sarah, E-commerce Operations Manager" at a $3M GMV Shopify Plus merchant: 80 disputes/month, 40 hours/month on manual evidence compilation, juggling order ops, logistics, customer service, and dispute management as a team of three. Built on a two-model Claude strategy — Haiku in production ($0.25/1M input tokens — latency and cost optimised for high-volume scoring), Sonnet for evaluation / development (prompt iteration and LLM-as-judge evals). n8n + Claude split architecture: n8n handles workflow orchestration, Claude handles the structured reasoning per stage.

Targeting the chargeback management market ($100B+ global chargeback volume per LexisNexis, 20–25% CAGR through 2027) where 60–80% of chargebacks are friendly fraud (not real fraud) — making evidence quality the deciding factor in dispute outcomes. B2B, sold to e-commerce business owners, finance leads, and operations heads at merchants. Purchasing trigger: 50+ disputes/month, spending 20+ hours/month on manual evidence compilation. Live prototype at dispute-iq.vercel.app.