Features
GRIFFO elevates the value of end-to-end, ‘democratized’ data/AI operations — from automated data check-in to reliable data governance and insightful data analytics — across any organization with AI-driven intelligence and automation. Embracing a no-code philosophy, GRIFFO fosters collaboration among extended AI teams, including business users, technical users, business analysts, and data scientists. Whether a company possesses in-house data/analytics expertise or not, GRIFFO ensures the capability to develop, deliver, and manage tailored AI solutions.
Automated Data Ingestion
Configure Once your Data Check-in Pipelines in the UI (without writing a single line of code) & Count on their On-Time Execution according to your Differentiated Needs.
Equipped with different batch, real-time and streaming methods, you can consistently onboard your own data and access external data complying with all popular formats, in the schedule you define.
End-to-End Data Interoperability
Unveil the Domain Knowledge behind your Data in a universal, flexible, and long-term sustainable manner.
Through AI-based support, you can semi-automatically map and transform your data according to the provisions of the appropriate domain-specific data model. The structure and the underlying semantics of your data become self-explanatory and suitable for (re-)use beyond the boundaries of a single department/group of users and/or the context for which they were originally created. Any need for updating/extending the underlying data models can be easily catered for through the UI, without requiring any development effort or affecting any data that have been already stored. (Bronze-level Data: Comprehensive Datasets)
Data Quality Enrichment
Accelerate the Τime and Diminish the Effort needed to Curate your Data in order to make them AI-Ready.
Through user-defined (manual) cleaning rules and advanced cleaning strategies, the labor-intensive data cleaning tasks are automated; any missing, incorrect, duplicate, irrelevant or outlier data are automatically detected and handled on-the-go, increasing the overall data quality and utility before they are stored. The degree of interventions in your data is continuously tracked through appropriate metrics and is visible at any moment in the UI (Silver-level Data: AI-ready Datasets)
Collaborative AI Pipeline Design
Unleash the Power of your Data through the 3-Dimensional Composition of Data Analytics Pipelines in the UI (without writing a single line of code), according to your Differentiated Needs.
From the graph-focused perspective, a data analytics pipeline can be visually and interactively created in different sophistication levels by:
- Efficiently manipulating the input data through >85 built-in and customizable data manipulation recipes (ranging from aggregations and math operations to time-series operations, join functions and create-dataframe options),
- Training, Applying and Evaluating high-performing machine-learning or deep-learning models (e.g. for any regression, classification, anomaly detection or clustering problem) through out-of-the-box support for the most popular ML libraries and frameworks (including scikit-learn, Spark MLlib, TensorFlow, Keras, XGboost, and statsmodels, among others)
- Including for-loops and control iterations based on different conditions.
In the table-focused perspective, the experimental results of the data analytics pipeline up to a selected step can be inspected in an interactive manner to correct any errors and handle any unexpected behaviours. The result-focused perspective allows you to configure the visualizations you prefer from a library of pre-built charts and gain additional visual insights on the results of your choice. (Gold-level Data: AI-insightful Datasets)
Effortless Transition of AI Pipelines to Production
Relish the automated deployment of the AI pipelines without any overhead or manual effort for adjustments or packaging prior to running them in production settings.
Featuring scheduled execution options (e.g. every day at 12:00), trigger-based execution options (e.g. when new input data are available) and external execution options (execution triggered via API by an external application), your AI pipelines run exactly when needed without wasting any resources. On-demand retrieval of the AI-ready data or the AI results can occur at any time from external applications/systems you use in your everyday operations (upon proper configuration in the UI).
Actionable Monitoring of all Data/AI Operations
Continuously observe your data check-in pipelines and your data analytics pipelines and track the health of their corresponding data/AI results.
With the help of custom monitoring views in the UI and user-defined, proactive alerts according to your needs, you obtain a 360-degree view of your data/AI operations in order to quickly troubleshoot the detected failures/incidents and efficiently address any underlying problem. Through our sophisticated observability features, your pipelines shall consistently perform with high reliability, your data/AI results will remain healthy and of high quality and your overall AI solutions will be robust.
Fine-grained Access Control
Put in place appropriate guard rails during your data-AI journey.
Each member of your extended AI teams remains in full control of the resources they own (both at the level of pipeline and asset) and may define appropriate access policies according to their preferences to permit seamless, yet “governed” collaboration privately (with other users, within their department or across different departments) or with the overall organization.
Deployment in any cloud provider
Avoidance of the cloud provider lock-in since GRIFFO can be deployed in the cloud service of your choice.
Lorem ipsum dolor sit amet, consectetuer adipiscing elit. Aenean commodo ligula eget dolor. Aenean massa. Cum sociis natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Donec quam felis, ultricies nec, pellentesque eu, pretium quis, sem. Nulla consequat massa quis enim. Donec pede justo, fringilla vel, aliquet nec, vulputate eget, arcu.