Design Architecture and Service Delivery
To provide simple access, speed and scalability we deploy models as an API at Phoenix Intelligence. The advantages of a solid micro-services design architecture are numerous.
Phoenix Intelligence Micro-Services
We’ve discovered that when dealing with data pipelines, the same sorts of challenges exist, two of which are data streaming and pre-processing, thanks to our work with customers in scoping and consulting. Phoenix Intelligence has created Micro-Services that solve some of these major concerns and can be implemented on any system Infrastructure.
Streaming Data
Install and manage Kafka pipes on Kubernetes to stream Data from any source to any public, private, or on-premises server. Data Streamer’s primary features include:
- The installation of a secure Kafka cluster.
- Configuration of a broker on a case-by-case basis.
- An SSL/TSL tunnel is used to encrypt communication.
- Data services are being monitored via.
- Dashboard for Prometheus.
- Fault-tolerant.
The sharpest minds in AI and Data Science at Phoenix Intelligence use tried-and-true best practices to ensure the success of AI initiatives.
Preprocessing of Data
On Kubernetes deploy and manage a Spark pipeline. Phoenix Intelligence has developed a Spark pipeline that can pre-process Data such as financial time-series, and deliver it to any public-private cloud or on-premises server.
The following are some of the most important features:
- Pipelines for scalable pre-processing.
- Popular Machine learning tools, such as Scala, Java, Python, and R have high-level APIs.
- Spark SQL for SQL and DataFrames as well as MLlib for Machine learning are higher-level tools.
- GraphX is a graph processing program.
- Spark Streaming is a stream processing framework.
- Fault-tolerant.
