This becomes one of the selling points of the project I was working on.

Even if the free version, you can do nearly whatever the hell you wanna do with the system. With small-size business, 100,000 requests per month is enough to give you life. There are a ton of library that you can integrate into, though I'm using NextJs and run into some troubles at first. The technical lead was basically always there to give me direction lol even though I was on my free plan.

May take you a little bit of coding experience if you want to integrate in react or nextjs. Also sometimes the connection is slow, it throws timeout exception (you have to manually set it)

It has been a great experience both from a product standpoint but also a customer service standpoint. [SENSITIVE CONTENT HIDDEN] has been extremely helpful in helping tonebase get set up.

1. The price point is unbeatable compared to other services 2. Its extremely easy to implement from a software engineering perspective 3. It has a robust set of features but they do not crowd the interface, the core concepts are very clear 4. The documentation was very clear as to how to implement the actual integration

Only would love a button to promote the data from staging db to production db. If recombee can add that to their product roadmap that would be extremely helpful imo.

Yeah, we have personal recommendations! (But we don't see any significant slift-ups in our orders.)

- It's very easy to use - Fast an good results - Very flexible

- No match score available - Expensive, if you use all features. Caused by per request pricing. - Client-Caching required to control costs

Best we could find that was easy(ish) to implement.

Nothing comes to mind at the moment, maybe responsiveness

The very powerful full-text multi entity search engine was the first feature we used in beta state for our customers but in my opinion the best feature is the more or less classic recommender engine service that drives 90% of the pages at some of our customers e-commerce solutions.

Like mentioned above the search engine is at Standard level(at the moment). But I don’t think that really is a „con“. They improve it and give more and more options by the time we are running our systems on that. So It stays stable and we and recombee can focus on new features.

The accuracy is what sets this recommendation engine apart. Recombee team also ensures that customer support is better than any other company.

None that come to my mind. We still need to test it with other languages though.

The integration with their API is very easy.

A data center option in South America would be very interesting.

It's a very nice choice that the small team could take advantages from your database immediately. No complicated algorithms, clean documents and fast support.

Recombee is super easy on integration, we took less than 2 days to make it work. And the result is acceptable. Its very good option when you are in a small team which can't afford a whole data-analyze team.

We can't find any additional solutions / suggestion to make the recommend result more accurate. The performance dashboard is a bit too simple.

We found both the product and the team at Recombee to be highly responsive and knowledgeable in their area of expertise, at this time (several months after launch) we still do not have an AI\ML team member and rely solely on their product and suport.

We create a new platform and wanted to have AI based recommendations for our users however we didn't have the time and resources to build the models and starting training them ourselves, the fact we could provide Recombee with meta data on our users and products and get going in a matter of minutes was exactly what we needed.

Nothing much, it provided us exactly with what we needed.

Excellent, I loved the user interface and it makes too many tasks easy for you. It is a product that can be used and based on obtaining follow-up statistics

It is different from other options on the market, its interface seems too understandable to me

Artificial Intelligence optimization is for experts, I think it should be more accessible. I mean hyperparameterization

We use a number of custom-built models to power our software, but chose to go with Recombee for our recommendation engine as it outperformed everything else we tried, and allowed us to launch much sooner. The docs are easy to consume and very comprehensive, the API is well-thought-out and supports a range of different use cases. It ended up saving us weeks of development time.

One neat addition would be the ability to see a confidence score on an item recommendation. We'd love to show a "Chance you'll like this" score.

The easy to use interface. It makes it really easy and fast to see what operations your users are doing.

I don't have a least favourite part of the software