AskAlibi has answered 99,000 product questions on Alibi's website
Dealers, installers, and prospects ask technical questions on alibisecurity.com and get answers grounded in Alibi's own documentation, 208 conversations a day.
The answers existed, getting to them quickly was the problem
Alibi's catalog spans cameras, recorders, accessories, compatibility information, application guidance, and competitive product comparisons, and the information customers needed already existed. It was just distributed across datasheets, product pages, and technical documentation. That creates the same problem for two very different visitors: a dealer or installer confirming compatibility mid-design, and a first-time prospect working out which products fit a requirement before talking to anyone. In both cases the answer exists somewhere in the documentation. The challenge is getting to it quickly.
Product knowledge became an answer layer visitors query directly
Alibi turned its product knowledge into an answer layer visitors can query directly. AskAlibi runs on Alibi's own specifications, application guidance, compatibility information, and cross-reference data against competing camera brands, and reasons over all of it together, domain-aware, rather than treating each document as a separate lookup. It reaches visitors through two experiences embedded on alibisecurity.com: AI-enhanced product search for direct questions, and a conversational assistant for the complex ones that span multiple documents.
An actively used part of the website, not a feature on it
AskAlibi has handled 99,323 conversations, roughly 208 a day across the deployment, on Alibi's public website with no login in the way. For companies weighing whether customers will keep using an AI product experience after launch, sustained volume at that scale is the signal: this is not a feature that exists on the site, it is how visitors find product information. The scope has grown too, from launch at GSX 2025 to a competitive cross-reference catalog covering competing security-hardware brands.
Two ways in, one knowledge layer
For direct product and specification questions, visitors search in natural language instead of leaning on filters, keywords, or individual datasheets. For the harder questions, the conversational experience retrieves across multiple pieces of Alibi documentation and returns a sourced answer. Both run on the same grounded product knowledge, so the answer does not depend on which door the visitor walked through.
Rapidflare made it easy for us to launch this project with minimal effort from our team.
George Farley, CMO, Alibi Security
Public traffic, every stage of the journey
AskAlibi is not restricted to a support portal, dealer login, or known customer account. It sits on Alibi’s public website, so the same product intelligence layer serves someone researching Alibi for the first time, an installer comparing products, and an experienced dealer hunting one specific technical answer. The architecture point matters beyond Alibi: companies do not necessarily need separate AI systems for product discovery and technical product questions. One knowledge layer can carry both.
A launch, then a widening scope
AskAlibi launched publicly at the GSX security industry conference in September 2025. Since then the scope has grown to include a competitive cross-reference catalog, so visitors can ask how Alibi’s products compare to competing security-hardware brands, not just look up Alibi’s own specs. Rapidflare has kept expanding the underlying product and documentation coverage as the catalog evolves, working with Alibi’s product and web teams to ingest updates and review answer quality, rather than treating launch as a one-time build.
How these numbers are measured
The figures here are cumulative product-usage totals pulled directly from Rapidflare’s live analytics for the Alibi deployment. Rapidflare does not currently have Alibi data connecting AskAlibi usage to support-ticket deflection, dealer onboarding time, pipeline, or revenue, so this case study reports only what we can observe directly: conversation volume, message volume, and deployment usage. If business-outcome data becomes available, it belongs alongside these numbers, not inferred from them.
“Rapidflare's AI agents consistently exceed the accuracy requirements we thought possible. They decode complex specifications and present them in simple, usable answers.”