Take GenAI from Idea to Product in Days
Easily discover Models in the registry. Test them manually with the embedded UI. Integrate the code in your service. Monitor status on UI.
FIND
Easily browse and search the over 1200+ Models in our registry. Test them manually with the embedded UI, tuned for each model.
Integrate your proprietary and in-house models with ease - they will also profit from our API driven run-on-demand platform.
Control which Models your audience has acess to with a simple configuration file - synced with GitOps, of course.
TEST
Test your models and prompts within the UI.
View the sensible defaults and tune your prompts.
Compare results across manual runs.
Upload files to secure object storage for processing.
USE
The AiFlow UI shows you both the Sync and Async calls to each API along with sensible defaults.
Use our OpenAPI compliant gateway to query models and integrate them with your applications.
The API also provides for emitting events to wire your results to downstream services
Integrate the code into your implementation specific services.
Wire them to into complex pipelines in minutes.
ORCHESTRATE
Create workflows with multiple service integrations easily.
Build, extend, change and deploy pipelines incrementally without disrupting running services.
Integrate with 1000s of Databases, Messaging Systems, APIs, Clouds, Kubernetes, ANYTHING with ease
Create parallel flows, AB Test models, Feed RAGs and more.
MONITOR.
View model prediction status and flows within UI, along with execution times, starts, ends, logs..
Be alerted on Service failure.
Temporarily de-activate a single service or maintain skip lists for multiple services.
Support service optimisation by keeping track of running averages of slowest services, most failing service etc.
Pragmatically Crafted
AIFLow is built of top of StreamZero's tried and tested Automation platform. Providing you an easy to leverage approach for enhancing your enterprise with AI.
Designed to Support Multiple Teams, Well Organized Code and Maximum Observability
The FX Hub Links Services Without Any Code Annotations and Provides Granular Information on Service Flows
Review and Manage Configurations with Ease
Schedule Runs with Well Known Cron Expressions. Monitor Runs on Dashboard
Monitor Runs and Flows with Ease. Integrated Metrics Collections and Analytics. Manual as well as Programattic Retries.
Scaled for Prime Time
Architected for Scale on Apache Kafka and a State of the Art Event-Driven Architecture
Feature Rich
Yet designed for everyday usability.
Our platforms are designed to serve as Enterprise Development Platforms with multiple teams and divergent use cases
Role based authentication, multi-project support, secrets management, aggregated logging, programmatic retries, automated and programatic re-prioritization, cross-network eventing, execution log archiving ...
All features engineered based on decades of project experience.
Powered by Kubernetes
All platform infrastructure runs on Kubernetes. Giving you the possibility to scale on demand. Each layer of the stack is designed to scale indpendently of the others.
Run Millions of Executions
Events are transported through Kafka and profit from the resilience of Kafka - Enabling event and execution throughputs of Millions of executions per day day in and day out. If the underlying infrastructure can support it.
Easily Scalable Executors
Executors run in easily scalable clusters which are further subdivided by priority lanes.
Serverless Executions
Allows services to run as launch-on-event containers, enabling dynamic scaling and cost-efficient execution without the need for dedicated server management.
Streaming Inputs
Add edge adapter to stream the input - Apache Kafka, Apache Pulsar, Apache Spark, AWS Event Bridge, Twitter ....
Database Inputs
Over 40 SQLAlchemy powered DB integrations. Google BigQuery, AWS RedShift, Apache Hive, Clickhouse, Postgres...
Storage Inputs
Amazon S3, HDFS, Minio, NFS, Google Cloud Storage, Azure Cloud Storage...
GitOps-Driven Deployment and Configuration
Employs a GitOps approach for deployment and configuration, leveraging Git repositories as the source of truth for system states, facilitating automated, traceable, and efficient change management.
Stream Aggregated Logs, Events and Notifications
All logs are streamed to Kafka just as we do with events. Purpose built adapters take care of feeding them into storage such as Elasticsearch, Splunk, Clickhouse, HDFS or AWS S3.
Monitoring Dashboards
Use the Management dashboard to view the health of your system. Identify slow running executions and fix them easily. Get notified on failure. View precise execution time of each step within a service. Identify new code running on platform and who esecuted them.
Trouble Ticketing Integrations
Auto-generate trouble tickets on failure. Integrate with your ticketing system of choice - after all it is just a service away.
Long-Term Log Storage Integration
Facilitates integration with platforms like Splunk and Elasticsearch for long-term log storage, allowing for advanced analysis and insights from historical data.
Role-Based Access Control
Efficiently manage security configurations by user role and sub-projects and with syncronisation to group managements in LDAP or ActiveDirectory.
Auditing
All actions are aggregated and ready to ingest into the audit system of choice - Elasticsearch, Splunk, ObjectStorage.
Authentication
Fully integrated with KeyCloak the de-facto standard in authentication gateways. Integrate with your enterprise authentication, easily.
Secrets Storage
In-built secrets storage to save your applications secrets. Support for secrets at service, project and platform level.
Environment Isolation
Each Service runs in it's own environment - with it's own dedicated context.
Selective Payload Encryption
Payload attributes can be encrypted such that only selected services can decrypt.
Service Collections / Projects
Groups services into distinct Projects. Each with it's own distinct team and roles. Allowing you to run multiple use cases with distinct teams or suppliers on a unified platform.
Micro-UIs for Business Users
Attach forms to services to trigger executions - so business users can trigger on demand executions. Forms are generated from JSON definitions. Create forms in minutes.
Project Specific Dashboards
Each project has it's own sets of dashoards for monitoring - enabling each team to monitor their own services.
API Gateway
Offers a unified entry point for all API calls, simplifying API management and monitoring.
Use Normal IDE
Use your normal IDE to create Services code.
Single Library Access to Platform Features
All platform features are available from a single powerful library - a low learning curve.
No Code Orchestration and Configuration
Orchestration and configuration are managed external to code and by each service. No complex DAG design.
Push to Deploy
Fully GitOps based. Push code to Git and Sync at a click of a button to deploy.
Advanced Features Made Simple
Complex features such as execution retries, state management and payload encryption are a single line of code away.
Specification First or Code First Documentation
Services and platform documentation are created on the fly by platform based on running services. But you can also build specification first with OpenAPI and AsyncAPI support for code generation.