The Conversational AI Technology Landscape: Version 5.0
This market map originated as chatbot development framework focussed research.
Subsequently it started following the expansion of all related technology into areas like Voicebots, testing, NLU tooling and more…
This is most probably the last time I will be updating this chart due to a number of market shifts and developments…and the chatbot / voicebot market being fragmented.
Traditionally the chatbot architecture and tooling were very much settled around four pillars; intents, entities, dialog flows and response messages.
This has been disrupted by the advent of voice and large language models.
The Voice Disruption
The market requirement to automate voice calls originating from a telephone call placed to a contact centre, necessitated chatbot vendors to branch out into voice.
Voice as a medium demands two pieces of highly specialised technology; Automatic Speech Recognition (ASR, Speech To Text, STT) and Speech Synthesis(Text To Speech, TTS). The voice focus also shifted away from dedicated devices like Google Home and Alexa, to automated telephone calls.
This change in focus added complexity in the form of higher value conversations, calls with higher consequences, longer conversations with more dialog turns and complex conversation elements like self-correction, background noise and more.
The Large Language Model (LLM) Disruption
Large Language Models also disrupted the ecosystem in two stages.
The first stage was during chatbot/voicebot development; starting with adding efficiency to the NLU development process in terms of generating training data for intents, detecting named entities, etc.
An obvious step was using LLMs for copy writing and generating and vetting responses.
This developed into the process of describing a flow, and the framework generating a flow, with variables, responses and more.
The introduction of LLMs at design time was a safe avenue in terms of the risk of customer facing aberrations or UX failures. It was also a way to mitigate cost and spending and not face the challenges of customer and PII data being sent into the cloud.
The second stage of LLM disruption was at run time, where LLMs were used for Search assistants, and highly contextual conversations via RAGimplementations.
Read more on LLM disruption and other factors in a follow-up post…
⭐️ Follow me on LinkedIn for updates on Large Language Models ⭐️
I’m currently the Chief Evangelist @ Kore AI. I explore & write about all things at the intersection of AI & language; ranging from LLMs, Chatbots, Voicebots, Development Frameworks, Data-Centric latent spaces & more.
