A review from a "not-so-technical" perspective
September 15, 2017

A review from a "not-so-technical" perspective

River Hain | TrustRadius Reviewer
Score 9 out of 10
Vetted Review
Verified User
Review Source

Overall Satisfaction with IBM Cloud PaaS (formerly IBM Bluemix - PaaS)

While I am sure there are other orgs/teams within the company that use Bluemix, I am unfamiliar with anybody else's usage outside that of my groups. We use Bluemix for several things, but our main use thus far has been for the Conversation Service. We use this service for the dialog design of our Virtual Agent. Additionally, we are using WKS (which I hear is now no longer a part of bluemix?) for a content annotation project with the end-goal of creating somewhat of a semantic network for our content library. The broader goal here is to improve our content's findability and relevance, so after WKS we will begin piloting WDS as a content retrieval service. That being said, our primary use-case is by far the Conversation Service for dialog design.
  • Intuitive user interface makes it easy for anyone to use, regardless of their professional background.
  • A lot of the services integrate well with external platforms, APIs, and programs, not just IBM services. A lot of the competitors in this space lack this ability.
  • Maybe it is just our contract in particular, but support and help is always made available.
  • Need: VISUALIZATION CAPABILITIES! Particularly with the Conversation Service.
  • Need: Annotation capabilities for dialog nodes in Conversation Service.
  • Need: Search/querying capabilities in Conversation Service
  • Need: Clearer documentation of the S2T service. I had to use a third party website for an understanding of how to use this.
  • Our Virtual Agent (designed with Watson Conversation Service) handles more than 1,000 conversations per day.
  • Since February 2017, our Virtual Agent (designed with Watson Conversation Service) has saved 222,363 minutes, or roughly 463 days, of customer support agent's time working on less complex Tier 1 issues, freeing up agent capacity to assist customers with more complex inquiries.
  • A lot of our work is still a WIP so it is hard to answer this.
Wasn't a huge consideration for me particularly, but I know my manager considered this. That being said, it makes me feel better about my work knowing we are operating on an open-source platform that helps move toward a world wherein almost anyone can deploy through the cloud and contribute to a universal project.
We aimed to create a cognitive and emotionally engaging virtual agent using IBM Bluemix Services. To do so, a level humanization is needed. Communication is a two-way street, and thus cognition is necessitated from both a recognition and response perspective. In other words we needed a set of services that would enable us to create a virtual agent who could understand natural language, cognitively compute the signals, and thus respond appropriately. We aimed to replace specific human tasks, so we needed to replicate the type and levels of human cognition needed for these tasks.
This is not really my area of expertise. I personally did not use Docker for any of our deployments, nor do I know if our Application Develops.. I wish I had more feedback for you here, but I don't. I know we just added a Linux instance, so I would guess Docker software was used for this?
Once again, I really wish I had some feedback for you here, but I don't. I don't even know if our application developers use it, but I see potential value in cost savings by using this, or a service like this, to avoid paying for the other resources/services/API calls we don't make/use during a session.
I have use EC2 and Microsoft's Azure. To me, both Azure and Bluemix were fantastic, but they each had some pros and cons. Azure had more services to offer, but their biggest flaw was in their inability to integrate and work with external platforms, APIs, Programs, etc.. Like Azure, Bluemix services work really well together, but unlike Azure they also integrated well with the other programs we were using for our work.
Well Suited:
- Development of information architecture/library. It enables better classification/taxonomy, leading to more intuitive findability.
- Dialog design and content retrieval for virtual agents. (e.g. a virtual agent whose content offerings are not hard-coded into the response fields, but instead require crawling/drawing from other pages/libraries)
Not Well Suited:
- Annotation/labeling/clustering of information that will be retrieved using a different search/query service.

IBM Cloud Foundry Feature Ratings

Workflow engine capability
Services-enabled integration
Development environment creation
Development environment replication
Issue monitoring and notification
Issue recovery
Upgrades and platform fixes