API monitoring and SLAs – Interview with APImetrics

September 21, 2017 Stephane Castellani

I’m very happy to welcome this week David O’Neill, CEO and Founder of APImetrics.

Stephane Castellani: Hi David, can you please present APImetrics Software in a few words?

David O’Neill: APImetrics was founded in 2014 in Seattle, and is currently Angel Funded with investment from Bain Capital Ventures and a number of individual investors. We have 5 employees and customers ranging from small security companies to large global banks.

SC: You are addressing the API monitoring market, can you explain us why this is so critical today to monitor API performance efficiently?

DO: APIs are rapidly becoming the life blood of the internet, as more and more services connect over APIs, businesses have become dependent on their working properly. Beyond that more companies are moving to business models where they are monetizing their APIs and they need to have a mutual way of agreeing what is acceptable performance and resolving disputes between the stakeholders.

SC: How did you come to the idea of grounding APImetrics?

DO: We had been working on the roll out of a new set of APIs for an international consortium of carriers. However, during the deployment phase we were having significant problems with the availability of the APIs. When we raised this with the provider of the API Gateway (not Axway!) we were told that there had been 100% uptime during the period we had trouble. So we built V1 to be able to prove things were working.

SC: Which are the different scenarios for API monitoring?

DO: Our customers fall into 2 camps:

  • Either they provide APIs and want to ensure that they are performing as expected.
  • OR they are consuming a service via API and are having problems that they  want to understanding better.

SC: Who are your customers? Which industries do you address and what are their specificities?

DO: We have an enormous range of customers, which itself is indicative of much penetration APIs have got. This includes banks, vehicle companies, security companies, marketing platforms and telecommunications businesses.

SC: Which channels do you sell your product through? Online, via sales teams, via partners?

DO: We sell online and through partners. For our enterprise customers we usually help with the set up phase to ensure they derive maximum benefits from our product.

SC: What’s your pricing model?

DO: We charge an annual fee based on the volume of monitoring you want to do.

SC: Can you share with us a recent customer success story, indicating the challenges they faced and the outcomes they got with your product?

DO: A customer, actually an Axway client, was having issues with the responsiveness of their APIs. They had extensive monitoring and testing with both New Relic and Selenium based testing, with Elastic Search and Splunk. With all these tools they were still unable to agree on the performance impacting issues.  Within a few hours of configuring APIs metrics they had traceable examples of API calls which had been slow and were able to find that the problem was related to a piece of legacy backend technology which they had not suspected would impact the APIs. Once they had fixed configuration, they were able to proceed.

SC: Which recommendations would you give to our readers who are about to deploy APIs?

DO: First, understand who your end user is. Where in the world are they and how does their infrastructure work? We see enormous differences between how a platform calling from the US works compares to Asia, we also see differences between the Cloud host like Google or Azure.

If you have a global customer base, can you have a distributed architecture? The difference between a call made in the US to a service hosted in Virginia and the same call made from Singapore can be 1-2 SECONDS. If your service is time critical or part of a sequence of calls then that can have a huge impact on users.

Finally, understand how consistent your results are. Be very nervous of your averages, instead look for the outlying results and the frequency of those outliers. Those are what will impact your users and lead to unhappy clients.:

SC: Can you share with us your product presentation video?

 

SC: Are there any other topics you would like our audience to be aware of about your company?

DO:  As more and more businesses become reliant on APIs either to provide or consume services the industry needs to start to take the question of Service Level Agreements more seriously.

  • How do you measure quality?
  • What does good mean?
  • How do different stakeholders measure the same thing?

We have found it very, very easy for operations teams to report what they can measure rather than what they SHOULD measure.

SC: Thank you David, this was an excellent overview of your expertise and the value you can bring to API providers and API users.

DO: Thank you Stephane.

 

The post API monitoring and SLAs – Interview with APImetrics appeared first on API Friends.

Previous Article
List of specifications for your API Management project (Part 1/2)
List of specifications for your API Management project (Part 1/2)

Succeeding with an API Management project requires good discipline and following three major steps: a good ...

Next Article
What are the different types of APIs?
What are the different types of APIs?

There isn’t just one type of APIs but actually three main types of APIs: Open APIs – aka Public APIs … More...