AIFA

The problem

500 million SMBs worldwide sell physical goods from catalogues. Their customers now expect to shop inside the channels they already use — Telegram, WhatsApp, Instagram — but building and maintaining a separate bot or storefront for each channel is too expensive and too technical for most small businesses.

Existing tools either work on a single channel only, or require months of custom development. The result: SMBs either miss the channel entirely, or pay a developer every time they want to expand.

Market size

• $290B — conversational commerce market by 2025, growing 25% YoY
• 2B+ — WhatsApp monthly active users; 500M+ business messages daily
• 800M+ — Telegram monthly active users
• 3B+ — Facebook monthly active users; 1B+ use Marketplace monthly
• 2B+ — Instagram monthly active users; 70% follow at least one business account
• Fashion, bookshops, florists, delis, photographers, cosmetics, clinics, hospitality, real estate — any business with a catalogue and repeat customers is a viable tenant

The solution

AIFA turns a business catalogue and domain knowledge into a 24/7 AI sales assistant, deployed simultaneously across web, Telegram, WhatsApp, and future channels — from a single engine, with a single setup.

• 🛍 AI-powered catalogue search and product recommendations
• ❤️ Wishlist and cart — unified across all channels
• 💳 Checkout with Stripe payments
• 📦 Order tracking and status updates
• 📬 Order notifications delivered to the seller in Telegram or WhatsApp
• 📐 Size recommendations and saved delivery address

Competitive landscape

Technical moat

A channel-agnostic hexagonal architecture means adding a new channel (WhatsApp, Instagram, Facebook) requires zero changes to the core engine — only a new adapter. Unified customer state (cart, wishlist, size, address) is maintained in a single store and reflected instantly across every platform. No data silos, no sync lag.

A multi-tenant RAG engine serves all businesses from shared infrastructure, keeping per-tenant compute cost low while maintaining full data isolation.

The architecture is designed for explosive, scalable growth — new tenants and their customers are onboarded without structural changes. Horizontal scaling is built in from day one: increased interaction volume is absorbed smoothly, with no degradation in response quality or latency.

Revenue model

• Setup + SaaS — one-time onboarding fee plus a monthly subscription per tenant, tiered by catalogue size and channel count.
• Full-cycle digital catalogue service — for businesses without an existing catalogue or web presence: end-to-end design, content production, and ongoing maintenance of a custom digital catalogue, delivered as a managed service.
• White-label licensing — agencies and platform resellers licensing the engine for their own clients.
• Revenue share — optional model for high-volume tenants: percentage of GMV processed through the platform.

Current state

Live working product. Web storefront and Telegram channel operational. Fashion vertical deployed end-to-end: search, cart, Stripe checkout, order tracking, seller notifications. WhatsApp integration in active development. All AI inference runs on self-hosted infrastructure — no dependency on third-party API quotas.

• ✅ Web + Telegram — live
• 🔜 WhatsApp — in development
• 📅 Facebook, Instagram — roadmap

The ask

Pre-orders open for early-access businesses. Seeking seed investment to accelerate channel expansion (WhatsApp, Instagram), grow the sales pipeline, and hire a second engineer. See the business page for the full seller proposition.