IBM Watson Discovery is an AI-powered search and text analytics platform that uncovers insights from large datasets. It is designed to enhance decision-making processes by streamlining data exploration and interpretation.
| Product | Mindshare (%) |
|---|---|
| IBM Watson Discovery | 3.2% |
| Elastic Search | 10.1% |
| OpenText Knowledge Discovery (IDOL) | 6.2% |
| Other | 80.5% |
IBM Watson Discovery taps into advanced natural language processing to help businesses analyze unstructured data with precision. It empowers users to extract meaningful insights quickly, driving efficiency and informed strategy development. By integrating seamlessly into workflows, it allows organizations to derive value from hidden patterns and trends, optimizing processes and understanding customer needs in a dynamic environment.
What are the key features of IBM Watson Discovery?Industries such as finance, healthcare, and retail implement IBM Watson Discovery to leverage AI-driven insights that tailor customer experiences, optimize operations, and enhance risk management. These sectors benefit from its ability to process vast amounts of unstructured data, drawing out actionable intelligence crucial for strategic development and operational refinement.
Prudential, Bradesco, Woodside
| Author info | Rating | Review Summary |
|---|---|---|
| Sales Director at Kukac | 4.0 | I use IBM Watson Discovery primarily with IBM Watson Assistant to enhance customer interactions by retrieving product information. Its conversational AI is valuable, but pricing needs improvement, as customers shifted due to the costlier version required for advanced use. |
| Research at a university with 1,001-5,000 employees | 4.0 | I use IBM Watson Discovery to assess model accuracy. Its Watson Suite integration and Assistant capabilities are valuable, though the newer version is harder to train. Setup is simple, it's stable, scalable, and support is good. |
| Research at a university with 1,001-5,000 employees | 4.0 | I value this solution for its help desk use, language support, and scalability, extending our service availability. Setup was straightforward. While flexible, it needs more effort for specific use cases and prebuilt models. I rate it 8/10. |
| Founding partner at Proxima TI | 3.5 | I value its entity extraction and content crawling. However, it demands significant memory (over 100GB for my 100GB index). While good for PoC, customer service slowness is a concern. I rate it 7/10. |
We use IBM Watson Discovery on some projects for our customers. We've used the product with IBM Watson Assistant to create solutions and answers about customer products.
Our primary use case for IBM Watson Discovery is understanding products by getting product information from some folders and documents. We have the IBM Watson Assistant in front, which gets answers or specific information from our records, and then it takes care of the conversation with customers, making it more human-like.
The most valuable feature of IBM Watson Discovery is testing, mainly because the product applies conversational AI, which means I can ask questions to get the information I want from a specific test area. Then, I'll receive a conversational answer from IBM Watson Discovery, even if I ask the question in a non-conversational format.
For general use cases, the product is okay, but for more technical requirements, that's still to be determined.
The pricing is an area for improvement in IBM Watson Discovery because the customer initially used the free version. Still, when he needed more questions and documents, he had to move to a different version, which was paid and cost $500 per month. That change in pricing made my company lose many customers.
We've been using IBM Watson Discovery since 2019.
The stability of IBM Watson Discovery is good, so it's a ten out of ten.
IBM Watson Discovery is a scalable product, so it's a ten.
I didn't speak with the technical support for IBM Watson Discovery because a different team communicates with support whenever there's a need.
The total time it takes to deploy IBM Watson Discovery depends on the documents you'll be working with. For example, I was in a situation where I was working with some painting files and folders for a painting store. The store had PDF documents, but the information was mixed up, so I had to treat the documents on IBM Watson Discovery, and discovering and understanding each PDF file took longer.
The process is more straightforward for plain documents, and you have to work with questions that will help IBM Watson Discovery understand the documents.
The time to deploy the product depends on the quantity and type of documents you'll be working with.
IBM Watson Discovery is an expensive product.
I have hands-on experience with IBM Watson Discovery. I'm a sales guy, so my product knowledge is for commercial purposes, which means I don't have in-depth knowledge. Still, I do know about IBM Watson Discovery functions, capabilities, and pricing.
My company is an IBM business partner. It's been a partner since 1994.
Deploying and maintaining IBM Watson Discovery requires one person from the Infrastructure team and two to three people from the Linguistics team who'll handle any questions about IBM Watson Assistant and do the training. Depending on the number of documents, there may be two to three more staff. It takes a minimum of one month to train for IBM Watson Discovery.
I will suggest IBM Watson Discovery to others looking into using it, but I will first show that its cost is high now, so the project must be well-designed. I'll also mention that another critical improvement area is that when you're installing IBM Watson Discovery, and you pause for some reason, or when you have to move it to a new installation, you have to do all the training again. You can't download, save, or move IBM Watson Discovery to another environment without redoing the training, which is another product con.
Overall, my rating for the product is eight out of ten, and I deducted two points for the two areas for improvement, the starting costs and the need to redo all the training when moving to another environment or when pausing the installation process for some reason. For example, I had a customer, an attorney, and because of changes in the company, the customer had to stop the IBM Watson Discovery contract. A year later, when the customer wanted to use the product again, all was lost. It could not be refreshed from a backup. There was no option to do that in IBM Watson Discovery.
We used IBM Watson Discovery to look at the different models that we have built and their performance accuracy.
The most valuable features of IBM Watson Discovery are the integration with the rest of the Watson Suite and the Watson Assistant capability. If you use Watson Assistant, the ability for it to be able to determine the accuracy of your voice models and your voice response systems is a benefit.
IBM Watson Discovery has moved to a newer version and it is very difficult to train the newer version. It's harder to train than the previous one. They do provide training but it is still a work in progress.
I have been using IBM Watson Discovery for approximately one year.
IBM Watson Discovery is stable.
The scalability of IBM Watson Discovery is very good.
We have three or four data scientists that use the solution.
The support from IBM Watson Discovery is good but could improve to make it great.
IBM Watson Discovery's initial setup is simple.
My advice to those wanting to implement IBM Watson Discovery is to take advantage of the trial and go explore it well in the 30 days before you adopt it. If you do adopt it, it works best within the Watson ecosystem.
I rate IBM Watson Discovery an eight out of ten.
We primarily use it for the help desk.
It has allowed us to provide longer service availability for support to users. Currently, so many people are working remotely or from home. Because of this solution, we didn't have to surge staffing to be able to meet the needs of so many users, especially when people are working longer hours and require more support, and are transitioning to a completely remote role.
Language support and the ability to build a natural language of speech recognition are the most valuable features.
There are probably other chatbots out there that were built for specific use cases and are easier to deploy than this. Having said that, Watson is way more flexible. While it may require a greater amount of effort, it is not substantially more than some of the other ones that are kind of prebuilt for a specific use case. It would be good to have more prebuilt and specific use cases and specific business models.
It can have better phone integration, even though I think that it is actually becoming less of an issue. Most people are online nowadays.
I have been using this solution for about nine months.
It has been solid.
It is very scalable. It is also a very broad and deep product. You can take it in a lot of different directions. It certainly could meet the needs of pretty much any business that is using this type of technology.
We develop and maintain a lot of solutions in-house. We have people from various lines of business who are actually making use of it. A lot of users of this solution are administrative sales staff. They probably are in the hundreds.
In terms of our plans to increase its usage, it is very hard to say at this point. It really depends on how much it is being used, but cost-wise, it is very reasonable, and it is based on API calls.
Their support has been good. We don't have any problems there. Their people are knowledgeable.
The initial setup was very straightforward. The deployment probably was a matter of weeks.
IBM supports a program, especially during COVID, to allow businesses to surge quickly. They were very quick, and it was a very positive experience.
For its deployment, we probably have two or three people.
Cost-wise, it is very reasonable because it is cloud-based.
If you can take advantage of some of the programs that are being offered by IBM right now to get people on Watson, you should do so because it's an investment part of IBM to do that. The biggest lesson that I have learned from using this solution is that it is opening up other use cases for this type of intelligence, chatbot type of applications, and other lines of business.
I would rate IBM Watson Discovery an eight out of ten. It is hard to give anything a ten. If it was easier to set up and had more prebuilt and specific use cases and specific business models, that would probably bring up the rating. The rating has nothing to do with its performance or its reliability. It has more to do with whatever efforts are necessary to build it for your particular business.
Being able to have some rules to extract the entities is valuable. The capability to crawl external sites and internal documents, and then draw internal information with external contents is also valuable.
It needs a lot of memory. Our index is very big. It is around 100 gigabytes. So, we need more than 100 gigabytes of memory to use Watson.
I have been using this solution since 2018.
We haven't deployed it to production yet. We are doing a proof of concept with some users. Currently, we have around 15 users, but we have plans to increase its usage. We currently have a solution for 400,000 users, and we want to replace this solution with Watson.
It's fine. Sometimes, they are kind of slow, which is okay when we are not in production, but in production, it is a risk for us.
It was simple because you get technical support from an IBM partner in Brazil.
I would rate it a seven out of ten. It's a good product.
