NLX

My main use case for Conversations by NLX is for multimodality experiences such as chat, voice, and text, as well as enterprise use cases with low code or some integrations, with different multilingual support. My role has a lot to do with testing different languages, so multilingual support is something I'm looking for.

I can give a specific example of how I used Conversations by NLX in one of my projects, where I'm particularly looking for differences between sarcasm, humor, and satire, and how it would respond to different things in different languages, as well as in different categories such as violent crimes and non-violent crimes. When we're doing adversarial or safety testing to check how strong the AI is, that's part of the red teaming effort that we are doing. We try different techniques, including different prompts which could trick the AI. The multilingual analysis of the prompt outputs in various categories is a critical aspect.

The best features that Conversations by NLX offers are voice and the multimodality experiences, especially for guided customer journeys across voice and digital channels.

Guided customer journeys across voice and digital channels are valuable for my team because we are a team of red teamers who do adversarial testing, and it really matters how it gives its output across different cultures, different languages, and different countries. It really has to stand out how the tools handle one channel well or if they're doing a great job connecting the experiences across different channels in one flow. Conversations by NLX positions voice to reduce friction and make the interaction feel much more seamless, especially for customers who really need step-by-step help.

Conversations by NLX has impacted my organization positively as a real differentiator for us because we do side-by-side testing for many AI models, and the multimodality experience in Conversations by NLX allows the customer to interact in more than one way during the same task, such as speaking while also seeing the prompts, choices, and confirmations. That has made a huge impact.

Conversations by NLX could improve on some advanced use cases, as they might need help with the technical setup and integration. That could be one downside, and sometimes it feels more enterprise-focused than a very beginner-friendly or lightweight tool. Some work in that area could be beneficial.

The pricing and implementation details of Conversations by NLX are not fully transparent upfront, so evaluation may require a sales conversation.

I have been using Conversations by NLX for almost a year.

Conversations by NLX is stable.

Conversations by NLX's scalability is pretty good, and they are very scalable.

Customer support for Conversations by NLX is good. Scalability is one of the main strengths, and if I've ever needed customer support, I've had a lot of support with customer interactions. I would rate the customer support for Conversations by NLX on a scale of one to ten as a ten.

I did previously use a different solution; however, this information is discreet, and I cannot share that. We did use a different solution at Google.

My experience with Conversations by NLX regarding pricing, setup cost, and licensing is that I'm not the one who is mainly in charge, but the pricing and implementation details were not very fully transparent upfront.

My company does not have a business relationship with Conversations by NLX other than being a customer; there is no relationship.

I have seen a return on investment with Conversations by NLX. We definitely saved money and time, but there were no staff that had to be let go because of Conversations by NLX. However, we could handle more projects because of it, so efficiency and time definitely improved.

The pricing and implementation details of Conversations by NLX are not fully transparent upfront, so evaluation may require a sales conversation.

Before choosing Conversations by NLX, I evaluated other options, and we evaluated many different options, even including Perplexity and Copilot.

The advice I would give to others looking into using Conversations by NLX is that they are pretty scalable, stable, accurate, relevant, and safe. I think everybody should use it.

If anybody needs a scalable, enterprise-grade conversational AI, they should go for Conversations by NLX. It looks pretty solid.

Conversations by NLX's governance and security look pretty secure, though I've never thought deeply about this parameter. It looks pretty solid on paper. They do call out the access control, the sensitive data handling, integrations, and runtime protections. For governance, I feel it's reasonably mature in the sense that the platform supports reviewability and operational oversight, but I would still want to verify practical details in the real deployment.

Regarding Conversations by NLX's AI capabilities, I feel they are accurate and reliable; almost nine out of ten times, it's accurate and pretty reliable, and the output is relevant, very consistent, and complete. I would rate this review overall as an eight out of ten.

Conversations by NLX is used for customers to get account details, card details, to block a card, or for loans, to get loan details, and to get a checkbook request or locate a bank branch.

We create a flow to do the task with Conversations by NLX by dragging and dropping and making a flow to accomplish this, and integrating APIs and REST APIs from the backend core banking APIs to authenticate the user or the customer. We can then proceed with the main task. For example, if a customer wants to block a debit card or wants any details, we can create a flow by creating the intents and entities. This defines what the user wants and what we are looking for. We can ask the required questions and create entities such as customer ID or registered mobile number as per the project requirements and specifics.

Conversations by NLX is user-friendly, easy to use, and environment-friendly, making it a good platform.

The drag-and-drop functionality of Conversations by NLX is easy to use. The environment is user-friendly, and we don't have to do much coding.

Conversations by NLX has helped us to assist our users and customers and develop our product for the contact center platforms.

I haven't used Conversations by NLX much, so I don't think there's an improvement needed.

Regarding Conversations by NLX's AI capabilities, the security is solid. I haven't done a deep dive into security for specific purposes, but it's secure and the authentication aspects were great.

I have been working with this solution for the past four years.

I am quite satisfied with the product we developed on Conversations by NLX.

We haven't had a partnership with another company. We are using a paid version.

Conversations by NLX is quite a good platform, and that's the main reason I have chosen it. It's easy to use, user-friendly, and you don't have to struggle to work on it. My overall rating for Conversations by NLX is seven to eight out of ten.

My main use case for Conversations by NLX is the chatbots, which I use most often.

A specific example of how I use the chatbots with Conversations by NLX is that the backend is working with NLX, and the best use case for us is the multilingual support that it provides.

What I appreciate most about the multilingual support is that while I think the accuracy is good, I am more inspired by the range of languages, and I feel it's quite deep, getting the context correctly.

Conversations by NLX has had a positive impact on my organization, and it's fairly a good review, I would say.

The positive impact of Conversations by NLX is that it saved time.

I think Conversations by NLX can be improved with more context or more training in that direction, as there can be some improvements.

I have been using Conversations by NLX for quite some time now, at least three to four months.

Regarding Conversations by NLX's AI capabilities, I think its governance and security are fairly okay, as it is binding to most of the rules, so I think it's okay.

I do not want to add anything else about the features; I think I'm satisfied.

I do not want to add more about the needed improvements; I think I'm satisfied.

I don't have an example of how much time Conversations by NLX has saved for my team.

I chose a rating of seven for Conversations by NLX because of the overall experience and the overall ease and the entire process.

I used Conversations by NLX once last year. The main use case for Conversations by NLX when I used it was to build a chatbot conversation bot for a BFSI client.

I built a chatbot wherein users, when called or using chat-based tools, get authenticated and raise queries such as raising a dispute regarding their balances, paying their account balances, or checking the account balances. The chatbot is an intelligent bot that gives users answers to their account balances after verification and asks them to pay the account balances immediately. If the balance is between $50 to $100, it provides one payment installment and a 10% instant discount. For $100 to $500, it offers four payment installments and a 12% discount, and it scales up to $10,000.

Conversations by NLX was pretty easy to use. The GUI was straightforward and anyone can use it without any prior knowledge. I am happy with it.

The best features Conversations by NLX offers are the NLP engine and the bot reading. What I appreciate most about the NLP and bot reading features is the bot training capabilities. I cannot compare it with any other tool currently, but it is straightforward and very easy to use, so I would rate it five out of five.

Conversations by NLX positively impacts my organization by helping in gathering customer inputs and providing resolution for their day-to-day conversations based on disputes, checking account balances, or paying account balances, thus reducing the average handling time. The average handling time was earlier around seven to nine minutes and has been reduced to five minutes.

One thing I would share regarding Conversations by NLX is the need to refine the self-service flows and integration capabilities for multi-modal usage and leveraging the analytics for personalization.

Regarding Conversations by NLX's AI capabilities, it has Intent automation that can trigger autonomous API calls. It can guarantee zero hallucination, prevent mixing topics, and wrap the conversation in a structured way.

For governance, I know it has RBAC (role-based access) and a virtual workspace. For security, it is compliant with data protection guardrails and enterprise-ready hosting, so it has all the security compatibilities required nowadays.

The accuracy and reliability of Conversations by NLX's output has core capabilities such as voice and multimodal AI, generative script workflow, and global reach. The output of the AI capabilities includes intent detection.

Conversations by NLX's intent detection and global reach has actionable outputs that can generate context-aware, cited references and can highlight information from large document sets, such as manuals or technical documents, directly to users.

The areas of capability improvement for Conversations by NLX are agentic AI and task automation, multi-modal UI, and intent classifications. These areas of improvement would enhance the voice and technology, expand the platform's ability to handle complex multi-step native calls, and expand support beyond native languages to add more languages.

Conversations by NLX is stable.

Conversations by NLX's scalability is pretty much scalable. It has a lot of speed to scale and can be scaled based on the number of chats or conversations it can handle. It has a seamless integration plugin to enterprise architecture such as Lex, Bedrock, or Contact Lens without disrupting existing backend solutions. This demonstrates the scalability of Conversations by NLX and it is beneficial that it can scale on its own based on requirements.

I do not require customer support for Conversations by NLX currently, but if I do require it or find out about the customer support in the future, I will definitely provide feedback.

I previously used Unifor for some clients and the Google Dialog bot with integration with Genesis. The choice depends on client requirements. All solutions are unique in their own ways and the selection would depend entirely on what the client's requirements are.

My experience with pricing, setup cost, and licensing for Conversations by NLX is that there is no upfront cost or setup cost. It has a feature-based pricing structure that begins with a free tier and scales depending on usage, whether they are using standard chat or multi-modal conversation. Depending on the capability being used, such as voice-plus technology, low code, omnichannel, or multi-language support, this is the best part of using Conversations by NLX and that is impressive.

AWS is used for deploying Conversations by NLX.

I have not seen a return on investment regarding Conversations by NLX yet. The client has not come back to me regarding the return on investment. However, it has definitely reduced the client's need to increase the headcount or the agent count after building the conversation AI tool or bot using Conversations by NLX analytics.

My experience with pricing, setup cost, and licensing for Conversations by NLX is that there is no upfront cost or setup cost. It has a feature-based pricing structure that begins with a free tier and scales depending on usage, whether they are using standard chat or multi-modal conversation. Depending on the capability being used, such as voice-plus technology, low code, omnichannel, or multi-language support, this is the best part of using Conversations by NLX and that is impressive.

I did not evaluate other options before choosing Conversations by NLX.

Conversations by NLX is designed for multimodal voice and digital self-service and can handle multilingual conversations. If they have similar usage, they can definitely consider Conversations by NLX because it can also leverage AI applications and you can bring your own LLM models to be used with Conversations by NLX. It can handle multilingual conversations in one frame of time and also has seamless channel deployment which can be tested in real time without wasting time or utilizing a native testing environment before going live. These are the tips and suggestions I would give to somebody considering using Conversations by NLX. I would rate this product a 7.5 out of 10.

I have not used Conversations by NLX in the production environment, but I have used it for proof of concept projects for three different customers in the UAT environment for the last six months. The main use case is for conversational AI where the call comes to our call routing solution, and from there, it gets transferred to Conversations by NLX to provide intelligent IVR treatment for customers. Conversations by NLX performs intent classification based on what the customer is saying, automatic speech recognition, and text-to-speech. It also conducts account lookup based on customer-provided information. A small amount of LLM or AI usage helps with better intent classification and FAQs. Once Conversations by NLX completes its processing, if it cannot solve the customer's issues, the call comes back to our call routing solution. Conversations by NLX also provides the most appropriate queue to which the call should be routed based on the current status of the account, and using that information, our call routing solution sends the call to that agent or queue.

What I like the most about Conversations by NLX is that the contact center industry, especially the legacy contact center industry, is very much accustomed to and comfortable using drag-and-drop blocks, connecting them, and writing the required business or application logic to serve business requirements. Conversations by NLX provides the same kind of web-based studio where you drag and drop blocks, connect them, and wherever required, you write business logic primarily using JavaScript. The studio is not only a place to develop your solution but also acts as a graphical representation of the overall solution. This is the first thing I appreciate about it, and it is what our customers appreciate about it, especially when compared with Amazon Lex, where they still have a graphical representation of things but not as comprehensive or anywhere close to Conversations by NLX.

The cost per call or per minute is relatively less. I cannot definitively compare it with the other two products that I have closely worked with, but the cost is relatively less compared to a couple of other competitors of Conversations by NLX that I worked with. The amount of coding required is less compared to other solutions, and while it still requires some amount of coding, it is relatively less when compared with other competitors. Integrating or building AI solutions or AI agents through Conversations by NLX is also easier compared to other competitors. These are the top four things that I appreciate or my customers appreciate about it.

I was not able to get approval to implement Conversations by NLX to more customers and business units because the cost is still high, especially when LLMs or AI get involved. That is one challenge regarding cost. Secondly, there are over half a dozen conversational AI providers who are doing exceptionally well in the US market as well as the Europe market. It is very competitive, so Conversations by NLX has to do a lot better, especially as I have difficulties promoting Conversations by NLX to some parts of European countries like Greece, Poland, and Austria. When it comes to European languages, they have to do a lot better. They are doing great with US English and Spanish, but compared with that part of Europe, it is a problem. Thirdly, it is hard for customers to open an account on the nlx.ai website and start building use cases. When I take Amazon or Google, it is relatively easy to create a contact center instance or conversational AI instance and start building use cases. Those kinds of things are not there in Conversations by NLX. I am also unable to reach NLX support to find answers for those things. Fourthly, since it has been acquired by Amazon, access to any new customers is completely blocked, and as of this month, it is simply not available since for the last several weeks. That is another big disadvantage.

I have been using Conversations by NLX for over six months.

Regarding stability, I would rate it ten out of ten.

For scalability, I would also rate it ten out of ten.

If you are a registered customer and need technical support regarding Conversations by NLX, it is easy. It is just that if someone is a random person trying to access nlx.ai to look around for training or access, technical support is not good. For registered customers, getting technical support is very easy. I would give a score of nine to the support on a scale from one to ten.

When I first started using Conversations by NLX, it was relatively easy for me to set up because I have worked on similar products such as Amazon Lex and Amelia. However, for any new customer, it is still difficult without proper training. Not just Conversations by NLX, if the customer is totally new to the conversational AI world, it is not easy. Much more training and access to build things for a reasonable cost is required.

Compared to other conversational AI solutions, Conversations by NLX requires very little maintenance. Once you are done building it, things like automatic speech recognition and text-to-speech going down or any latency issues are extremely rare. With that said, maintenance is relatively very little. I would give this review a rating of nine out of ten overall.

I have been using Conversations by NLX professionally for four years.

The best features Conversations by NLX offers include flexibility and the ability to loop back to the main question. I appreciate that it asks, "Is there anything else I can do for you?" so I do not need to initiate the same conversation again and go through the same flow because, as a user, I do not know the flow. If it helps me go back to previous stages of the conversation, that helps significantly, whereas a one-directional path makes it really hard for users to navigate.

Conversations by NLX positively impacts my organization by using an LLM to help us develop this type of tooling. However, I feel that it is also pushing the problem forward. Implementation is cheap, but review and quality are expensive. Thus, it pushes the problem to a further stage of the standard development cycle.

Improvements for Conversations by NLX can be made by focusing on context. The more data you push, the more personalized and problem-solving it becomes. For instance, using the phone assistant starts with a zero stage, knowing only my phone number, but it should also save some data. When starting the conversation, I expect this assistant to already know things about me. Data protection rules exist, but I want to feed it as much information as possible so the conversation flows better. It should at least know my job, name, age, and perhaps previous history of conversations, so I do not have to explain the whole history repeatedly.

For the fintech, I have been working for six months, and for the healthcare, it was around four years.

In my experience, Conversations by NLX is not stable, and I have not used it.

I have not switched from a different solution before this; I am still using the old one.

Regarding outcomes from Conversations by NLX, the implementation is probably cheaper, around fifty to sixty percent for many companies using this type of practice. This allows us to bring new tooling to the market faster and experiment quicker. In one day, during a hackathon, we can do a lot and showcase it. However, while this implementation saves time, it has pushed the quality issue forward. Making it production-ready becomes harder because dumping a lot of code does not necessarily speed up the job. Many people can work on languages and systems they do not know, allowing them to learn and implement faster, but I would still be cautious about the end stage, which is quality.

I have not seen a return on investment with similar systems since I found that usually the application and user experience are better rather than saving time or money.

Before choosing Conversations by NLX or similar systems, I did evaluate options, including a German company.

My main use case for Conversations by NLX is building one, and I am also a user of the same application, which is called the phone assistant by Doctorlib.

I use Conversations by NLX to communicate with my doctor's practice by calling the practice, where an AI phone assistant picks it up and asks what I need. I explain my needs, and then it goes into one of the flows such as appointment booking or prescription, and I can automatically book an appointment on the phone without using anything else.

As a user, it feels as though Conversations by NLX is trying to be smart, but it is sometimes hard, and I might need human interaction for this type of thing, especially for healthcare. When it involves something personal, I would prefer to have a personal human touch. AI systems are getting better and more personal, but some users still do not prefer talking to an AI system at all.

My application was a phone one, but Conversations by NLX can also be adapted to a web application. Usually, it is a chat interface to talk through, but I think it needs a revamp of how we interact with Conversations by NLX systems. I prefer it to be all the way through a phone call, not a chat interface because I do not want to write; I just want to talk or visualize it rather than reading the whole text.

Concerning Conversations by NLX's AI capabilities, governance and security are significant topics. Data can be shared, but it all goes back to security. Securing the models and having guardrails is really important because input and security posture are crucial regarding the attack surface. The attack surface is extensive, as it only accepts text and can do anything with it. Therefore, you need to control what you are feeding it and limit its access and manipulations.

My advice for others looking into using Conversations by NLX is simple: try it. It is not complicated. I would rate this review nine out of ten.