Endor Development Update — June 2019
Hello from the Endor team!
Endor Protocol v.1 was released for public use this past April. Since then, we’ve been going to great lengths to share with you, our community, the ins and outs of the business and product development. Through this development update you will get a brief recap of our progress over the last month and an idea of where we’re headed. It’s always someplace fascinating, we’re sure you’ll enjoy the ride.
THE LAW OF THE LAND
Great products are defined by many parameters. The ones you focus on, reflect in many ways the attitude of the people behind the technology. At Endor, we’ve dedicated the past few months to improving scalability, stability, and above all usability, because we want to encourage not only growth, but mainly large-scale user adoption as a necessary step towards truly democratizing AI.
SCALABILITY
For any product to grow, it must be scalable. For it to grow while continuously improving the service level it provides, scalability must be built into its core. That’s why we fixed 5 major infrastructure issues and can now run 10 times more predictions at any given time, at the same cost. Additionally, we began to adopt the AWS Glue service (learn more about it here: https://aws.amazon.com/glue/), working closely with the AWS team on further improving our infrastructure, to support rapid onboarding of new data at 5 times the previous rate, while keeping human intervention to a minimum and avoiding process mistakes. This tremendous progress takes us closer to our external data onboarding milestone, which is planned for later this year, and will allow other data providers to join our ecosystem.
STABILITY
While scalability enables growth, stability ensures a high service level, and that is something we take very seriously. But being truly dedicated to stability in a hyperelastic cloud environment such as ours, with thousands of AWS spot servers crunching raw data entails many challenges. We are continuously working with the AWS team to resolve these challenges, mainly related to the EMR service (learn more about it here: https://docs.aws.amazon.com/emr/latest/ManagementGuide/emr-what-is-emr.html). In order to compute the massive amount of data we rely on within short timeframes, we use top-notch distributed processing approaches. As these are relatively new services, we are constantly working on resolving edge cases (3 over the past sprints), which allow us to quickly upscale using auto-scaling capabilities, while maintaining the cost and stability of the infrastructure.
USABILITY
We’ve already stated the importance of scalability and stability in the ecosystem’s growth strategy, but the truth is, customer adoption is the most crucial part of any customer-facing product. We began by developing a very tech-savvy user interface, as our first milestone was setting up the right APIs for our partners, including Algowave and others. In our upcoming releases, we plan to focus on increasing product usability for the less technical users, so anyone can use our platform.
The current interface requires the end user to upload lists of ‘addresses of interest’, in order to generate predictions:
The new flow will support predefined, easily accessible queries that will require no special data preparation, allowing users to simply select the prediction settings and run:
Instead of uploading a file, they will click on the ‘List to Rank’ button, and select a file to work off from a predefined list:
Users will easily adjust the prediction settings to suit their query:
That’s it for now… We’re happy to have launched this communication channel and are committed to publishing development news once a month. So, stay tuned, there’s a lot more where this came from, the pipeline is exploding and we’re moving forward fast.
Onwards and upwards,
The Endor Team
EDR is listed on the following exchanges：
Upbit , Bitrrex ，Hotbit , BitForex ，Houbi korea , CoinTiger ，Bilaxy，Trade.io , Coinall , KuCoin , Idex , DigiFinex , P2PB2B , OAX , CoinBene , LATOKEN , BitMart , Coinbit , ABCC .
