What is our primary use case?
We have built an Enterprise Chatbot Foundation Platform. We built a B2C Chatbot Solution for retailers users and HR chatbots for employee Self Service functions.
These functions included:
For the B2C chatbot:
- The key objective was to bring down call center traffic with the chatbot
- Weekly 25,000 unique users interact with bots every week
- 78% of traffic rate Chatbot as 5-Stars
- Delivering a delightful CX to end-users
- Feedback and Recommendation captured
- Product containers and accessories
- Sales (coupons and discounts)
- Order status and notification
- Recipe related queries
- Live agent feature
- Order tracking number
- Call back IVR integrations
-
CRM integrated curated offerings
For the HR Self Service chatbot:
- HR Chatbot hosted on SharePoint
- Offers self-service chatbot
- HR help and questions
- Benefits and well-being
- Health policies and claims
- 30+ entities (20 sub-menus: visible)
- Feedback (capture feedback)
- Mood elevator (9 moods)
- Service ticket creations
- Huge impact during the Covid-19 pandemic
- Addressed 5000+ requests
- Covid-19 related FAQ and policies
How has it helped my organization?
The AI/analytics powered B2C Virtual Digital Assistant (VDA) solution fits perfectly into their B2C website and provides complete support automation for both their employees as well as their customers. Available 24/7, this solution can scale at peak times during festivals which makes them an excellent support resource.
This solution is designed for delivering an enhanced user experience to end-users and offers faster onboarding of any enterprise user. The struggle of customers/prospects due to the lack of an interactive channel to get answers to their queries was addressed with this solution. It also enables loyalty program features based on the occasions. It also helped in bringing down 18% of helpdesk calls and 40% of email inquiries.
The HR Self Service chatbot helped a great deal during Covid-19. Related FAQs addressed updated policies, benefits, and wellness and delivered a seamless experience to employees.
What is most valuable?
Amazon Lex offers fully automatic speech recognition and natural language understanding technologies to generate a speech-language understanding system. Its bots provide the capacity for multi-turn conversations. It helps context control natively so that you can manage the context immediately without the need for custom code.
It provides two types of prompts: confirmation prompts and error-handling prompts
It natively enables integration with AWS Lambda for fact retrieval, updates, and commercial enterprise logic execution. It allows you to simply publish your bot to chat services precisely from the Amazon Lex console, eliminating multi-platform development efforts.
What needs improvement?
AWS Lex has a complex web integration.
It has fewer deployment channels. Lex is not multilingual and supports only English.
The preparation of the data set is complicated; the utterances and entities mapping is somewhat critical.
To support other multi-language functionality with Amazon Lex, you’ll need to combine the use of other services including Amazon Comprehend and Amazon Translate to support non-English languages.
The maximum duration of speech input is 15 seconds. Amazon Lex bots require an internet connection and cannot be accessed locally.
For how long have I used the solution?
We have been using the Amazon Lex chatbot service for three years for our B2C chatbot and HR Self Service Chatbot Solutions.
What do I think about the stability of the solution?
It's a very stable platform and has had zero platform-related issues in the last three years.
What do I think about the scalability of the solution?
The solution offers high performance. We had zero performance-related issues in the last three years. It is the most stable and is a robust platform that is easily scalable to different regions globally with minimal configuration.
How are customer service and support?
So far, our experience with AWS Customer Service and their support team has been great. I'd give them five stars. They have been very responsive and the CSM team helps customers in product adoption and reviews everything free of cost.
How would you rate customer service and support?
Which solution did I use previously and why did I switch?
We used the Oracle Digital Assistant Chatbot solution, however, we switched due to the lack of features and surprise price hikes made by Oracle.
How was the initial setup?
The initial setup is very straightforward and easy to provision.
The AWS documentation is 100% up to the mark, and the Customer Success team will make sure the customer is onboarded swiftly. They work to resolve any technical queries very quickly. Sample Bots are provided and are available on GitHub, which accelerates the onboarding time for new customers.
What about the implementation team?
The customer implemented it through a systems integrator based out of India. The expertise of the SI has been outstanding and exceeds expectations as they helped in identifying the platform and persona-driven use cases. They helped with creating business cases, taking business approval, and implementing an AI-assisted bot in the iterative Digital Factory Model successfully.
What was our ROI?
The solution gives us a very high ROI. The chatbot platform cost has been recovered in just a few months. The platform itself costs us less than $5000 annually and is taking care of 1 million requests from end-users. On top of that, the AI manages the chatbot and has managed to make online sales worth $75,000 in just six months.
What's my experience with pricing, setup cost, and licensing?
The setup is very easy; you can provision a Lex environment in 30 minutes. The pricing was also the cheapest.
With Amazon Lex, you pay for what you use. It is charged based on the number of voice or text requests processed by your bot. At $0.00075 per text request and $0.004 per voice request, you can use Amazon Lex without spending a dime with the free tier version.
Which other solutions did I evaluate?
We evaluated AWS Lex, Google Dialog, and Azure platforms for chatbot functionality.
What other advice do I have?
AWS Lex, Lambda, Dynamo GB, API Gateway were used to build an integrated solution with on-Premise mission-critical applications.
Which deployment model are you using for this solution?
Public Cloud
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Amazon Web Services (AWS)
Disclosure: I am a real user, and this review is based on my own experience and opinions.
