Platform
An easy and affordable way to make your business smarter
Somma.ai is the first multi-cloud and no-code AI platform fully designed built in Latin America. With it, you can build unlimited AI engines without writing a single line of code.
If building AI applications on Somma.ai is easy, it's even easier to hire it!
Somma.ai has a Platform as a Service (PaaS) business model.
The processing infrastructure in the cloud is a customer choice and its costs are transferred transparently to our customers.
To run a solution on Somma.ai, one must prepare the data, assemble the analytical flow (engine or application) from native or custom components using the drag and drop menu, or use an existing flow, choose a cloud service of your choice and voilá – your solution is ready to use. All this is performed using Somma Studio, our web development environment that runs on any browser of your choice.
Platform architecture
- Within Somma Studio (1), the user builds the solution, defines the infrastructure where it will be processed (3), defines the origin and destination of the results (4), setup visualization schemes (5), and connect the platform with various data sources and outputs (6).
- Processing is started, monitored through color code showing the status of each component and through reports (logs) and stopped via Somma Studio (1).
- Processing can also be controlled through other applications using specific API and can be scheduled at your convenience.
- The user has full control over the origin and destination of his data and results, which cannot be accessed by third parties, and the server that will perform the processing, chosen by the user via Somma Studio (1), will be for his exclusive use.
- The user can visualize the data and results via dashboards (5) or use the analytical solution to automate processes (6).
Somma.ai
processing
flow
The technologies that support SOMMA
Big Data/HPC
Apache Spark, Apache Tika e Apache Arrow.
Native Storage (datasets and files): MongoDB.
MongoDB
Machine Learning e Deep Learning
Spark ML, Spark MLib, Scikit-learn, Gensim, NLTK, Tensor Flow, Keras, Pandas e outros
Connections with other databases
MySQL, PostgreSQL, Mongo DB, Mongo DB Atlas, S3, Google Drive, Apache Kafka and others.
