Thoughtful Chatbot Series: Passage AI

Episode 7: Building better chatbots with Passage AI

Hosts: Fahad Shoukat and Andrew Wolfe from Skiplist.

Today's guests: Mitul Tiwari, Founder/CTO, and Muckai Girish, VP Business Development, Passage AI

oMdmD9S5_400x400.jpg

Chatbots are here to stay. There is no doubting the tremendous potential chatbots hold in kicking customer service into another gear.

Once we get past the potential use cases for a technology such as chatbots, the next challenge is around implementation. Which platform? Google, Alexa, Cortana, Messenger or the 20+ more platforms out there. How about all of them?

We haven’t even touched integrations, security, privacy, and future proofing. It can be a daunting task, and we are just talking about chatbots here. Wait, I want my chatbot to be intelligent as well. I heard about this thing called AI and machine learning.

On the surface, chatbots sound like a good idea. However, actual implementation is another animal.

Fortunately, companies like Passage AI are making chatbots easier to build and implement. As we think about thoughtful software at Skiplist, Passage AI is tackling thoughtful bot building.

Andrew and I were fortunate enough to speak to Mitul Tiwari (CTO and Co-Founder) and Muckai Girish (VP of Business Development) from Passage AI.

A fantastic discussion around platforms, security, privacy, implementation, and trends around chatbots. You’ll learn more about the Passage AI platform, and they shared their expert insight into building a thoughtful bot.

If you are considering a chatbot, listen to this episode first!

Cheers,

Fahad Shoukat

Outline

Passage AI

Founded in 2016

VC-backed, $10M

  • Differentiators: Vertical-focus and use cases, end to end bot building platform, support of many languages and NLP implementations, supports messaging apps and tools.

  • Speed of deployment and accuracy are also strong characteristics of Passage AI's tech.

    • 100% necessary for bot success

  • Using off-the-shelf tools to build proprietary tech for the backend of Passage AI

  • Great use-cases for Passage AI

    • Customer service automation

    • Retail and telecomm

    • 24/7 response

    • Knowledge base search and automation

Chatbot Adoption (9:31)

9:31- End user awareness around bots?

  • Most customers want to be transparent with their end users about the fact that they're speaking with a chatbot.

  • Expectation is that the end user would be handed off to a live agent if necessary.

  • Uses sentiment-change detection in the chatbots to trigger hand-off to live agent sometimes.

  • We're far more comfortable with these digital tools, socially, now than ever before.

  • Passage AI is constantly adding support on social and messaging and voice platforms as well as some mobile apps.

What Differentiates Passage AI from Existing Platforms? (19:58)

19:58- For instance using instead of and Alexa skill?

  • Customers who've developed a Skill then want to scale it to other platforms. Can't do it.

  • Trying to help customers do efficient work: build once and deploy nearly everywhere.

  • Lots of market fragmentation - Passage AI is an easy way to integrate functionality across all these different hardware platforms.

  • Chatbots are also assisting professionals at the customer level, not just the end user level.

  • Chatbots aren't actually replacing humans at all; not even close.

Bot Security Concerns (26:12)

26:12- Security and bots

  • Stripping out confidential information

  • Process messages in real time to make this filtering work at the front end

  • All requests come in HTTPS. Data encrypted at REST.

  • Compliance with financial and SEC regs + GDPR compliance.

  • Following security best practices.

Internal vs External Adoption of Chatbots (30:05)

30:05- adoption for internal teams (e.g. IT, operations etc)?

  • Yes, really big adoption for internal teams

  • Password resets for IT, for instance

  • Expenses and travel policy implementation.

  • Insight into sourcing departments and MSA agreements at a company

  • This is most important because the smallest support teams are the ones supporting internal processes.

Protection Against Biases (32:45)

32:45- How do you approach identifying and protecting against common biases?

  • Introducing bias-training data.

  • Balance of training data across use-cases.

  • Using analytics to identify new bias issues if/when they come up and retrain

Trends for 2019 around Chatbots (38:17)

38:17- What's around the corner for chatbots?

  • 2018 - conversational interfaces

  • 2019 - contact center solutions

  • Chatbot is going to be a critical component of a call center. Chatbots will literally be handling the first part of incoming calls.

  • Customer service automation is still just ramping up and will continue to grow.

  • More services from tech companies means much more incoming support calls. Chatbots will be critical to handling these increased loads.


Mitul Tiwari is the CTO and Co-founder of Passage.AI. His expertise lies in building data-driven products using AI, Machine Learning and big data technologies. Previously he was head of People You May Know and Growth Relevance at LinkedIn, where he led technical innovations in large-scale social recommender systems. Prior to that, he worked at Kosmix (now Walmart Labs) on web-scale document and query categorization, and its applications. He earned his PhD in Computer Science from the University of Texas at Austin and his undergraduate degree from the Indian Institute of Technology, Bombay. He has also co-authored more than twenty publications in top conferences such as KDD, WWW, RecSys, VLDB, SIGIR, CIKM, and SPAA.

Girish Muckai is VP of Business Development & Strategy at Passage AI (www.passage.ai). Over his 20+ year career in high tech, he has worked with Juniper Networks, Reliance Jio, Entropic Communications, ARRIS Group and SBC Labs, helping bring broadband and wireless communications solutions to market.