Kaspr Documentation

Kaspr is a Kubernetes-native stream processing framework that dramatically simplifies building real-time, event-driven applications on Apache Kafka. Write your stream processing logic in Python using simple YAML configurations instead of complex distributed systems code.

Built as Kubernetes Custom Resource Definitions (CRDs), Kaspr provides enterprise-ready features like horizontal scaling, authentication, persistent storage, and monitoring out of the box. With Kafka as its only dependency, Kaspr offers the flexibility of custom stream processing without the complexity of traditional frameworks.

Key Features

• Declarative Configuration - Define entire applications using YAML, no complex DSL required
• Python Processing - Write business logic in Python with familiar libraries and patterns
• Stateful Operations - Built-in support for tables, joins, aggregations, and lookups
• REST API Integration - Expose web interfaces with KasprWebView for real-time data access
• Production Ready - Enterprise features including RBAC, resource limits, and fault tolerance
• Kubernetes Native - Seamless integration with existing Kubernetes infrastructure

What You Can Build

• Real-time Data Pipelines - Process and transform streaming data with complex business logic
• Event-Driven Microservices - Build loosely-coupled services that react to Kafka events
• Stream Analytics - Calculate real-time metrics, aggregations, and rolling computations
• Data Enrichment Services - Augment events with external API calls and cached lookups
• Business Rules Engines - Apply dynamic rules and routing logic to streaming data
• Notification Systems - Fan-out events to multiple channels with preferences and rate limiting

Core Resources

Kaspr applications use three main Kubernetes resources:

• KasprApp - Defines infrastructure, scaling, and Kafka connectivity
• KasprAgent - Processes streams with Python logic and routing rules
• KasprTable - Provides persistent key-value storage for stateful operations
• KasprWebView - Exposes Web APIs for real-time data access and interaction

Quick Start Guide

Simple Example

Here’s a minimal Kaspr application that processes user events:

apiVersion: kaspr.io/v1alpha1
kind: KasprAgent
metadata:
  name: user-event-processor
spec:
  input:
    topic:
      name: user-events
  output:
    topics:
      - name: processed-events
  processors:
    pipeline:
      - enrich-event
    operations:
      - name: enrich-event
        map:
          python: |
            def enrich_event(value):
                # Add processing timestamp
                value["processed_at"] = datetime.utcnow().isoformat()
                # Normalize user ID
                value["user_id"] = value["user_id"].lower()
                return value

Ready to build real-time applications with Kaspr? Start with the Introduction to learn the fundamentals.