Photo by Vernon Raineil Cenzon on Unsplash

With OneReach AI, If You Can Think It, You Can Most Probably Do It

Are we seeing the emergence of a Conversational AI orchestration engine and marketplace, a Twilio of Language Technology?

Cobus Greyling
5 min readJul 26, 2022



I again had the privilege to sit down with Robb Wilson and Jordan Ratner for over an hour to talk about what Conversational AI should be and to what extent human langauge is involved.

I often have these conversations with founders regarding their product and platform. What made my conversation with Robb and Jordan different from other conversations, was that to a large degree the platform called OneReach AI drifted into the background. They where more interested in talking about the experience and level of functionality customers and makers deserve.

Naval often talks about escaping competition through authenticity, I think the first seed regarding this notion of being authentic was planted by Peter Thiel. Is the approach of OneReach AI to act as an aggregator and orchestrator to establish highly contextual and sequenced user experiences, a way of escaping the competition?

I have argued before that OneReach AI is not a chatbot / voicebot development framework, as defined in the generalised image below.

OneReach AI can best be described as an orchestration engine acting as a single front-door for customers and not as a Conversational AI framework per se.

For instance, the OneReach AI design canvas is not a conversation flow designer, or flow node manager. It is an orchestration canvas, where multiple processes can be kicked off and managed in parallel. Micro-bots can be orchestrated for multidimensional customer service.

Why have all the chatbots gone back to live in IVR-land?

~ Robb Wilson

The OneReach AI Approach

A few key considerations regarding OneReach AI

  • OneReach AI endeavours to move away from a centralised logic engine, towards a more event driven approach.
  • Does user interactions always have to revolve around language? What about interactions where no conversations are involved?
  • Why does a select group of people in an organisation have the privilege to create conversations and flows?
  • Why can’t employees develop user experience sequences and submit their skills?
  • Should a decentralised bot building approach not be followed?
  • Data, information and procedures are used across an organisation, should skills not be created at each of those points, by the people who know best?

A sequenced flow of user interactions do not necessarily have to include human langauge.

~ Robb Wilson

  • The goal of OneReach AI is to create highly contextual, multi-turn experiences and interactions.
  • Conversations are not just language…a customer interaction can be a contextual, time-sequenced flow of events.

Gartner Research found that 42% of companies think their biggest challenge in adopting AI is not understanding the use cases.

~ Gartner

Patterns & Templates Patterns & Templates

Back to the idea of aggregation and users having access to a library of functionality. According to OneReach AI there are “tons” of patterns and templates, with more being added on a continuous basis.

Characteristics of patterns and templates are:

  • A lot of problems have already been solved.
  • These can be seen as tools you need to create solutions without reinventing the wheel.
  • Patterns and templates incorporate best practices to create predesigned steps, skills, flows, views and cards.
  • Specific steps all the way through to full solutions that you can add to your applications.

Why do you have to choose a specific vendor, why not choose them all?

~ Robb Wilson

Again, these tools go back to the idea of a decentralised bot building process, knowledge and data are disparate and are managed in different areas of the organisation.

Why can’t people in the organisation build their own interfaces, the ones who know the data and customer needs. Employees should be able to create a sequenced flow or functions which they can combine in a skill.

And subsequently submit skills and contribute to the bot with granular decision making. In turn these skills need to be orchestrated into a unified conversational interface.


The key consideration with OneReach AI will most probably be cost, and to what extent OneReach AI consultancy will have to be made use of.

Another consideration is the cost incurred from third party products accessed via OneReach AI. For instance, all the external services accessed and orchestrated from OneReach AI are costs above and beyond the cost making use of OneReach AI.



Cobus Greyling

I explore and write about all things at the intersection of AI & language; LLMs/NLP/NLU, Chat/Voicebots, CCAI.