Ecommerce AI assistant

An Ecommerce AI Assistant for Product Discovery, Recommendations, and Support

One catalog-aware assistant across Shopify storefront chat and WhatsApp. Cybergine gives ecommerce teams grounded product recommendations, compatibility answers, and support replies from their own catalogue data. Built by Eugene Zubanov, formerly an ML engineer at Expedia, Decathlon, and Product Madness, it works across Shopify storefront chat and WhatsApp from a single dashboard.

Product recommendations grounded in catalogue data
Compatibility questions, asked first
Support across storefront & WhatsApp
Grounded in your catalogue
Storefront + WhatsApp
Your storefront
● Catalogue-aware
Will this rack fit my road bike?
A couple of questions first: is it a 700c frame, and does it have rack mounts on the rear stays? That decides which option fits.
🚲
Pannier Rack (700c)
In catalogue · Mount required
🧰
Seatpost Rack
No mounts · Universal fit
Ask about fit, compatibility, returns…
Grounded answers from catalogue data WhatsApp + Shopify web chat No developer required One dashboard
Grounded, not generic

What an ecommerce AI assistant should handle

Cybergine is built for the commercial and support moments that matter before and after purchase. The assistant retrieves relevant product records before answering, so shopper replies stay tied to catalogue data.

01 / DISCOVER

Product Discovery and Recommendations

The assistant retrieves relevant product records from your catalogue before it answers, so recommendations stay tied to what you actually sell instead of generic AI patterns.

“What waterproof bags do you sell?”
Ridge Pack 22L · Dry Creek Tote · pulled from your catalogue
02 / FIT

Compatibility and Fit Questions

Customers ask whether parts fit together, whether a product works with their setup, or which option matches their use case. Cybergine asks the clarifying questions needed to narrow down the right answer.

“Will this part work with my setup?”
Clarifying questions first · then a grounded recommendation
03 / SUPPORT

Support Answers That Stay Product-Aware

Returns, shipping, sizing, and product-care questions still need grounded answers. The assistant can combine policy content with catalogue context so the reply stays useful to a shopper, not just technically correct.

“Can this jacket be returned after a rainy hike?”
● Return policy · Product care · Storm Shell record

Built by an ML Engineer, Not a Product Team

I built this because enterprise vendors charge more per year than most small merchants earn per month, for systems that cannot hold a coherent product conversation. I spent years building product search at Expedia, Decathlon, and Product Madness, systems that handle millions of products and millions of queries. Cybergine is that architecture, sized for small ecommerce teams.

No investors means no pressure to extract maximum value from every renewal. No product team means I talk to merchants directly and ship fixes in days, not quarters. If you email me, I reply.

Eugene Zubanov · Founder, Cybergine

Works Where Your Customers Are

Use this page if you need one product-aware assistant across channels. Cybergine runs as a WhatsApp chatbot for ecommerce stores and as a Shopify AI chatbot for storefront support, both connected to the same catalogue and dashboard. Customers who prefer WhatsApp get WhatsApp. Customers browsing your Shopify store get the storefront widget.

WhatsApp Business

Cart recovery, product questions, and order updates on WhatsApp: messaging-first setup connected to the same assistant knowledge.

See the WhatsApp Business Chatbot

Shopify web chat

Product recommendations and compatibility checking in the storefront widget. Start here if you want the support-first Shopify page.

See the Shopify AI Chatbot

Single dashboard

All conversations, all assistants, one place. Every channel reads from the same catalogue so answers stay consistent.

Start free

Common Questions

Does it work for my product category?
Yes - the assistant is trained on your catalogue via RAG, not on a predefined product category. Whether you sell bicycles, electrical parts, or fashion, the assistant learns from your product data.
How does it avoid hallucinating product information?
The assistant answers from your catalogue using retrieval-augmented generation (RAG). It retrieves actual product records and generates answers from that data. It cannot invent specs or prices that are not in your catalogue.
What does it cost?
Pricing depends on the channels and setup you need. Start free and Cybergine will recommend the right starting plan for your store.

Ready to add an ecommerce AI assistant to your store?

Start with catalogue-grounded product answers across your storefront widget and supported messaging channels.

Grounded answers · Storefront + WhatsApp · One dashboard