Coveo AI-Relevance Platform: Enterprise AI Search and Answers

Coveo's main use case revolves around content indexing and search integration, primarily when working with Sitecore as a CMS implementation.

In our Sitecore CMS implementation, we have very large content volumes, and Coveo helped us significantly with relevance through the use of indexing, query pipelines, and security features. Relevance was the most valuable feature for us. We were able to determine what content was shown to users and which was not, so Coveo provided substantial value from a relevance perspective.

From a business perspective, Coveo's biggest value is having more control over how users discover content, including relevance, ranking, personalization, and analytics, which can help improve the search experience without requiring a developer to change application code for every small adjustment. The exact value obviously depends on how well the organization configures and uses those capabilities.

Before moving to Coveo, we had metrics of user clicks around two million, but after we went live with Coveo, user clicks and all view analytics jumped up over 3.5 million, which was a significant upgrade after implementing Coveo.

In my view, the best feature that Coveo offers is relevance, which is one of the strongest areas. You have more control than you would typically get with a basic search implementation, particularly through query pipelines and ranking configuration. The important thing is that good relevance still requires proper configuration and understanding of your content, so I would not expect the platform to solve every relevance problem automatically.

I find query pipelines particularly useful because they give you a way to influence search behavior based on the context of the query without having to pull that logic into your application. For example, you can apply different ranking on query behavior depending on the situation, which provides a good separation between the application and search configuration.

Recently, I have been exposed to working with Coveo Atomic while working on the headless Sitecore implementation, and I appreciate the Atomic library because it provides reusable search components, so you don't have to build every part of the search interface from scratch. It makes the front-end integration faster while still allowing the developer to customize the experience when needed.

The main area I would improve is the learning curve because there are many concepts and configuration options, and understanding how indexing, query pipelines, security, and all those pieces work together requires experience. Coveo's onboarding experience could be simpler, especially for developers using the platform for the first time, and I believe that improvement on the documentation is also needed because while Coveo has extensive documentation, it is sometimes difficult to find exactly what you need.

As a developer, when trying to find something regarding how to do indexing, I have to read the documentation, but if there were screenshots showing how it can be done, it would make a developer's life simpler. Additionally, including code examples would greatly enhance the experience for developers.

I have been using Coveo mainly in enterprise search implementation, particularly in Sitecore environments, so I have been working with Coveo for the last five to six years.

From an enterprise perspective, scalability is one reason we consider a platform Coveo because we don't have to build and maintain the underlying search infrastructure ourselves, and the platform is designed for large content environments. My direct experience is more focused on implementation and integration, but I haven't seen scalability become a major limitation.

I recommend starting with search requirements rather than starting with product features. Understand your content sources, security model, and relevance.

So far, I haven't touched Coveo's AI-powered solutions, so I may not be able to comment on that because my experience is mostly with integrating Coveo, but not with the AI-powered solution models. I haven't used those AI features much, so I cannot provide insight on the accuracy and reliability of output.

I would rate Coveo an eight out of ten. I choose this rating because the platform is powerful, has strong relevance, and enterprise capabilities, but I deduct a couple of points mainly due to the learning curve and some areas that can take time to understand when implementing Coveo for the first time.

My main use case for Coveo is the search page. I use Coveo for my search page by creating sources in Coveo cloud organization and then providing it with data from our website and then displaying that search result on the search page.

Coveo has impacted my organization positively, as the search has been very good; the search terms and fuzzy search have all become very effective. I don't have particular data regarding specific outcomes or metrics at this time.

One thing that stands out to me about using Coveo for this process is that, as it is a cloud platform, it is very easy to integrate with your website as compared to other search options available.

The best features Coveo offers are the machine learning models and Coveo API. What I like about the machine learning models is that they are very easy to integrate; you just need to create some settings in the cloud organization, and then when you start using the search, it automatically learns from those searches and provides auto-suggestion in the search box.

Regarding Coveo's AI capabilities, I think its governance and security are good; as it is a cloud-based platform, the security is already very good as it uses SSO or login features. The accuracy and reliability of output from Coveo comes when the machine learning models learn from the searches that are done on your live site, so it becomes better over time, and I think those are reliable.

In terms of integration capabilities with other platforms or tools I use, I think it is pretty good; integration is straightforward. I have done it with Sitecore as well, and it provides out-of-the-box packages to integrate it, so it is very easy. I would describe the user interface and overall user experience with Coveo as very good.

I think Coveo can be improved by providing a sandbox for developers, even if you are not part of an organization, and the API configurations can be made a bit easier because there are a lot of options right now which confuses users. If I had to rate Coveo's flexibility regarding customization for my specific needs, I think it would be six out of ten.

I have been using Coveo for two years.

I think Coveo is stable.

I think Coveo's scalability is good.

The customer support for Coveo is very good.

I have used Solr before Coveo, and we switched to Coveo to utilize its AI capabilities.

Coveo is deployed in my organization as a SaaS-based software, so you don't need to deploy it; it comes by default on a cloud and it basically will provide you an organization so you can log in and then work with Coveo.

I can't say I have seen a return on investment with Coveo; I don't have these metrics.

I was not part of the pricing and licensing experience as it is handled by my organization, so I'm not aware.

Before choosing Coveo, we did not evaluate other options.

I would rate Coveo an eight out of ten. I chose eight out of ten because, as this is a cloud-based platform, there are some limitations, but overall it's a good tool.

Coveo already has a very good e-learning platform with multiple courses to learn and certifications to achieve, so it is very good in terms of the learning curve for new users or developers working with it. The documentation and resources available for Coveo are helpful, and even if you are a new learner, you can go through the documentation and it covers everything related to Coveo.

As far as I have used it, Coveo's performance and speed with large data sets is pretty good. I would advise others looking into using Coveo that it's a great product; just start looking into the documentation and utilize it well so your search page becomes better.

I used Coveo for content indexing and content search from AEM, and I have used it for multiple languages. The content was available in English language and Chinese language, so our requirement was to index the content and then search the content from Coveo.

Coveo has a push API which indexes the content with a certain structure and JSON payload for the page properties that exist in AEM pages. For example, a page title, page description, page tags, and metadata properties from the page are indexed in Coveo with the help of Coveo push API. There are certain request headers such as UTF-8 encoding. We wrote an HTTP client which makes calls to Coveo push API to index the content. To search the content, we used the Coveo endpoint and the page ID as the path, basically the page path as a resource to pull the content from Coveo.

We used Coveo for multiple languages with English language content and Chinese language content. We faced challenges initially because we had not added UTF encoding in the header, so the Chinese content was not properly indexed and instead showed question marks. We added the header with UTF-8 encoding, and after invoking the same API, we were successfully able to index the Chinese content. We fixed this issue and successfully indexed and fetched the Chinese content from Coveo.

Coveo is an enterprise-grade search and index platform that can be integrated with any CMS platform; in our case, it was Adobe Experience Manager. Since it is API-based, Coveo can also be used in other CMS platforms.

We did not face challenges in the integration process with Adobe Experience Manager specifically. We went to Coveo platform, reviewed the API integration methods, and learned how to pass the key and authentication methods in the API. We configured the OSGi configurations, which is typically how AEM supports any kind of configurations. We used the OSGi configurations in AEM and configured these API endpoints and the authentication methods, OAuth tokens, and other security measures. We invoked that API in our code base; in our code, we have OSGi services and Sling Servlets which are invoking Coveo's APIs.

Coveo is scalable and I have not seen any latency issue. The content immediately reflects the moment you index it, and in a couple of minutes, it is reflecting on the Coveo server. It is easy to use, easy to navigate, and you can easily find your content which is indexed on Coveo platform. Coveo is a specialized platform in the field of indexing and searching and is a known search platform. We seamlessly integrated and benefited, especially because indexing the content was taking too much time when we were indexing to other platforms such as Solr. However, in Coveo, we have not faced any latency issue and we were quickly able to index the content and search the content.

Coveo is easy to integrate and it hardly takes a few lines of code that we need to write. There are not many complex configurations we have to write for integration. A small configuration for the API endpoint, the API payload, and the authentication methods allows us to write a few lines of code for an HTTP client to invoke these APIs. It is easy to integrate and we can easily maintain it. If there are any changes in the future, we can easily make these changes.

Coveo can improve by supporting a few more languages. For example, we have tested with English and Chinese and a bit of French. I think Coveo can come up with indexing and searching with non-native languages apart from these known languages.

When we have to scroll down to search for particular pages, maybe they can provide easier navigation and links to find content, either with the help of page paths or another method. We have seen that if we have multiple pages or many pages published to Coveo to index, then it is time-consuming to scroll down and see the content, especially if it is indexed properly.

I have used Coveo for more than two years.

Coveo is deployed in our organization on a public cloud with API integrations and authentication methods, and we call this publicly hosted Coveo platform on the cloud.

We have used AWS, if I remember correctly, because it was a long time ago, almost more than two to three years back now.

There was a separate team actually accessing Coveo console and they were aware of that; we were using those APIs and the configurations as a vendor.

If others are looking into using Coveo, I would advise that if they have a use case for scalable solutions for enterprise-wide search capability or indexing content, then Coveo is the right choice.

I would rate Coveo a seven out of ten.

My main use case for Coveo is the search functionality that it provides on our website, which is heavily integrated with our content management system, enabling indexing of our content coming directly out of Sitecore, helping visitors discover content.

In the commercial real estate domain, we operate applications that allow listing agents or property agents to list properties for sale or lease. Once they create a property listing, they push out the data to the content management system, which immediately triggers Coveo APIs to index all the pushed content, populating the search UI that visitors use to search and discover listings and other content.

Coveo's primary usage is how it integrates with our marketing technology platform through its search functionality.

One of the best features that Coveo offers is the AI-enabled query suggestions feature, which looks at user query patterns and suggests relevant queries when users revisit the site, driving engagement.

AI-enabled query suggestions help users return more frequently and build trust in the search capabilities our site offers, allowing us to drive content towards users based on popular queries. Coveo offers a lot of analytics, including reporting on the most used queries and search relevance, which helped us consolidate various search hubs to improve end-user experience.

Coveo has helped build trust among users, although we still have a lot to improve for the overall search experience. We measure trust or improvement in user experience through engagement metrics, looking at website visit frequency, users' success in finding what they're looking for, click-through rates, and how those convert into leads.

Coveo could be improved by simplifying its configuration and setup, as it has a lot of dependencies on Coveo expertise, making it less intuitive.

I have been using Coveo for the last one and a half years.

Coveo is stable; the indexing is reliable, and we can see the fallouts after indexing runs.

I have not seen Coveo being scaled significantly during the last one and a half years, with only a few configurational tweaks made.

Customer support has been adequate; we do not involve them too much.

This is my first experience with a search platform, and I have not used a different solution previously.

In a general sense, we are not seeing a remarkable return on investment because maintaining Coveo for our search functionality is costly, and we have to balance investments in search versus other areas.

I rate it at six mainly due to the complexity of setting up Coveo independently and its considerable expense as a product.

Those were not things I was personally involved in, but I heard feedback about how expensive it was to use Coveo, particularly regarding API calls.

Coveo is deployed in a private cloud in our organization. I am not really sure specifically which cloud provider that is for our private cloud deployment. Anyone using Coveo should be aware that customizations will involve more work whenever changes or improvements are made. I would rate Coveo a six out of ten based on my experience.