#IAConf Wrap-up: Bridging Chasms in IA Ecosystem

(written in collaboration with Amy Stapleton)

picThe combined Intelligent Assistants and Intelligent Authentication Conferences (#IAConf) have come to a close. Reflecting the raised profile for bots, intelligent assistance (IA) and artificial reality (AI), it was our largest conference yet, both in terms of sessions, sponsoring companies and attendees. Insights that were shared from the podium and in the hallways centered on the importance of both intelligent assistance and authentication for improving digital user experience, especially around customer care.

Collectively, the themes under discussion exposed a set of dichotomies, both for core technologies and their capabilities, when deployed for real-world use cases. Many of the core technologies are regarded as “just emerging” and considered by many to be nascent, even though they are rooted in decades of research, development and productization and have, in fact, been field proven in implementations that span contact centers, Web sites, interactive voice response systems, mobile apps, not to mention “The Cloud.”

The contrast was made most clear during a panel on the final day of the conference. Technology investors Phil Libin of General Catalyst, Sarah Guo of Greylock Capital, and Joshua Kauffman of Wisdom described the near-incredible promise of messaging-based bots and conversational UIs while confirming that we are in the very earliest days of providing promising implementations and use cases.

They agreed that developers and implementers, alike, are largely exploring how the new user interaction paradigm will play out, with Phil Libin providing the most vivid analogy by asserting, first, that successful bots, by definition, will be much “better than humans” at carrying out their assigned tasks. He further asserted that it won’t be long before “the cream rises to the top” and a meritocracy of top bots will emerge, reflecting the popularity and earned reputation of select bot “personalities”. By their actions, people will determine which messaging platforms become dominant, and how people and brands interacting with one another.

While this august group of early-stage investors see a world in which the game has just started, several speakers presented case studies that served testaments to the maturity, reliability and scalability of existing intelligent assistance resources. For instance, Jessie Blocker of Xapo, a rapidly growing company that claims to provide “the world’s most secure bitcoin wallet”, showed a compelling example of how smart self-service technology works. To onboard users and popularize its service, Xapo fields a high number of user inquiries with a very small customer support staff. As a young company, they chose to meet user demands by implementing Inbenta’s intelligent assistant platform, which enables customers to ask questions using natural language. Implementing this proven self-service technology has enabled Xapo to keep pace with customer growth while keeping operational costs in check.

What works for small start-ups has also proven out at large, transformational telcos, as demonstrated by Sarah Bramwell of TalkTalk, one of the UK’s largest providers of TV, broadband, and mobile services. She showed how the company leveraged both natural language technologies and voice biometrics to improve customer experience and retention during the course of its “transformation” to highly diversified, largely digital service provider.

TalkTalk implemented natural language understanding into their IVR platform with the result that customers are routed much more easily and quickly to the right call agent. They also pioneered voice biometrics, enabling customers to authenticate with voice when calling into TalkTalk. By authenticating, customers experience seamless routing to the agents and the answers they need and significant reductions in overall call times.

Scott Bair of Nationwide demonstrated how the insurance provider is using technology from Flamingo.ai to guide prospects through the process of selecting insurance that meets their needs. Flamingo provides an intelligent “sales” assistant that interacts with the prospect via a conversational interface, collecting the required information to select an appropriate policy and provide premium quotes. Nationwide has seen the intelligent assistant drive up sales numbers and continues to expand the use of the technology.

Each of these real life use cases proves that intelligence assistant and intelligent authentication solutions are available today and offer significant benefits to companies and their customers. This is a vivid contrast to the “state-of-the-art” in IA when Opus Research convened the first #IAConf in 2014. The core technologies were of interest mostly to early adopters. Now, with the Facebook-infused excitement around conversational UIs and messaging bots, almost every Fortune 5000 company understands the opportunities and is working on their “bot strategy.”

While bots that interact with users on messaging platforms are just getting started, intelligent assistant solutions based on natural language have already proven their tremendous value. Both technologies will continue to grow, with customers ultimately reaping the benefits. Discussions at #IAConf defined several contrasting visions that must be resolved (or at least mitigated) as the ecosystem evolves. Here is a brief list:

  1. If there are millions of special purpose, task oriented bots affiliated with various messaging platforms, how will individuals discover and choose which ones have merit and can do the best job on for them?
  2. In a world where individuals have their own “metabots” (meaning a single bot that operates on their behalf), what standards and protocols will be required to enable that individual to use his or her own words to express intent, find things, make purchases or simply carry out conversations that involve other bots, friends or other people (E.g. subject matter experts).
  3. By the way, how will those Bots and Metabots (like Amazon’s Alexa) identify and authenticate individuals in order to offer personalized services? Can continuous authentication take place to support such services across devices and channels both online and offline?
  4. On the flip side of the continuum, How are companies leveraging and augmenting existing investment in Interactive Voice Response systems, Contact Centers, CRM, Analytics, Knowledge Management, Interactive Voice Response to support both bots and IAs?
  5. What will be the business justifications for moving to new conversational models? Are we just building more sophisticated or conversational ways to deliver advertising? Will enterprises invest in the technologies solely to replace “expensive” staff in contact centers and elsewhere?

The refresh cycle for new technologies and platforms is accelerating and more gaps (emerging chasms) will arise as both developers and implementers bring new bots and IAs into the world. Collective efforts and constant conversations among researchers, developers, implementers and investors are called for. We, at Opus Research, hope to serve as facilitators. The conferences offer an opportunity for face-to-face meetings and contemporaneous discussions. Between events, we invite you to join our “Intelligent Assistants Developers and Implementers” group on LinkedIn to stay in the loop.



Categories: Conversational Intelligence, Intelligent Assistants, Intelligent Authentication, Articles

Tags: , , ,

2 replies

  1. Your observations about the breathless enthusiasm of VCs gazing on a new crock of gold versus battled hardened customer service operations getting on with it is well made.
    This is a market bonded by common technology yet divided by perception. I attended a digital agency sponsored bot meet up last week in London. Over 100 in attendance. The context was bots, AI and Messaging.
    That was both the context and assumption for the whole audience. Whereas a classic Inbenta use case would look quite different.
    These differing views on the backstory and history of the market will make the interoperability issues you raise even more of a challenge. Time to reach for the bot phone?

  2. Nice wordplay Martin! I think the next step in building bridges across various chasms is to “focus on the data” or knowledge management resources that are required to provide accurate/consistent/context-aware responses regardless of the point of ingress (bot, IA, chatbox…).