About eXvisory

eXvisory bot in a suit


It's time to change how companies support their products.

Customers have problems with products. That’s natural. But modern product support is terrible. Who doesn't dread calling telephone help lines? Hanging on hold for ages to talk to out-sourced support staff who know less about the problem than you do and can’t (or won't) deviate from apology-laden scripts?

Is asking your customers to search through megabytes of complaints and misinformation on self-help forums really the best answer?

The problem is that providing product support is time-consuming and expensive, especially for complex products sold to thousands of consumers. It's mission impossible. So, unfortunately, the limit of most companies' ambition is to be no worse at support than their competitors.

That’s what eXvisory aims to change.


Artificial Intelligence (AI) is hot again, thanks to advances in AI algorithms combined with Internet-scale data, compute and connectivity.

eXvisory uses a type of AI called deep logic networks (imagine IF...THEN statements crossed with neural networks in a shady Oxford UK biolab) to model the systematic trouble-shooting of human experts.

Customers interact with eXvisory support AIs via a familiar Q&A chat interface. Scroll up, drill down, revisit questions and explanations and share and collaborate on your ‘eXvisory’ with friends, family or colleagues (or even support humans).

CONVENIENT self-paced online troubleshooting. Start and stop whenever you like. No apologies. No flattery. No dropped calls. Get access to the best support expertise ... every time.

It’s so cool it's almost fun.


A big problem with logic AIs was that they needed specialised boffins to program them (in weird languages like LISP and PROLOG) and the boffins had to pester human experts to extract the expertise that was keeping the human experts in a job. eXvisory gets round this problem (and others) by imposing a novel application-specific structure onto its deep logic networks. But what does that mean?

Building old-school logic AIs is like hand crafting 1000-piece jigsaw puzzles from a tree. Building an eXvisory AI is like solving a shop-bought jigsaw. You still need to figure out where all the pieces go, but it’s not that hard. And the jigsaw pieces refer to articles from existing support knowledge bases. All you have to do is fit them together into an eXvisory AI deep logic network, using our user-friendly visual editors.

No boffins required.