ALGINTE

Manage Apache Kafka — and build stream processors visually, without a Java project.

Self-hosted, one container, free to run. Topics, consumer groups, connectors and schemas — plus a canvas that deploys real Kafka Streams topologies next to your cluster.

Download See the Streams builder

Build Kafka Streams topologies visually

Most Kafka UIs let you watch your cluster. Alginte also lets you build on it: compose a stream processor on a canvas, write the transformations as expressions, and run it — in minutes, without starting a Java project.

The Alginte topology canvas: a word-count stream with source, flatMap, groupBy, count and sink nodes, and a node's expression editor open — A word-count topology on the canvas, with the node palette and an expression editor open — click to view full size.

Draw the topology

Drag sources, processors, and sinks onto a canvas and connect them. The full operator set is there — map, filter, flatMap, branch, merge, group, aggregate, reduce, count, windows, joins, cogroup.

Write the logic inline

Transformations are SpEL expressions with a type-aware editor: completions from the key/value serdes at that point in the graph, live validation, and per-operator variables like #aggValue.

Deploy without a build pipeline

Submit and the topology runs inside Alginte, next to your cluster — no Java project, no jar to ship, no CI round-trip to test an idea.

Keep the design as JSON

Download any topology — configuration and canvas layout — and upload it back to edit or redeploy. Version it, share it, hand it over.

Deployed streams run in your instance and resume from Kafka-side state, so a redeploy picks up where processing left off. Full walkthrough and three importable examples in the Streams builder guide.

And everything you expect from a Kafka console

Topics & messages

Browse records point-in-time or live-tail them. Seek by newest, oldest, offset, or timestamp; decode keys and values as String, numbers, JSON, Avro, or Protobuf; produce test messages from a drawer without leaving the page.

Consumer groups

See members, per-partition committed and log-end offsets, and lag at a glance. Reset a partition offset to replay or skip — with the guardrail Kafka requires: only while the group is idle.

Schema Registry

Register Avro, JSON Schema, and Protobuf subjects with the naming strategy you use, diff any two versions side by side, and set compatibility per subject.

Kafka Connect

Create connectors through a form generated from the plugin’s own config — validated against the worker as you type — then watch tasks and restart just the failed ones.

Brokers & ksqlDB

Inspect cluster nodes, their configuration, and which partitions live where. Run ksqlDB statements and pull queries from a built-in editor when a ksqlDB server is configured.

MCP server 0.6.0

Point an MCP client at Alginte and let it read your cluster — topics, groups, schemas, connectors, sampled messages — through read-only, annotated tools.

The Alginte overview dashboard with state cards for brokers, topics, consumer groups, streams, Connect and Schema Registry — Cluster health at a glance — every card links into its area.

A consumer group's offsets tab showing committed and latest offsets with lag per partition — Per-partition lag, and offset reset when a group is idle.

Why teams run it

Runs next to your cluster

One container or one JAR, self-hosted. It binds to localhost by default, and no data ever leaves your infrastructure — there is no vendor backend in the path.

Free to run, commercially and in production

The Community edition needs no licence key, no registration, and never expires. Air-gapped installs keep working.

Works with your Kafka

Smoke-tested each release against Apache Kafka 3.7 through 4.2, Confluent Platform, and Redpanda — cluster management and deployed stream topologies alike. It connects as a client; nothing is installed on your brokers.

Starts in one command

No database to provision, no operator to install, no sidecars. Point it at your brokers and it is up.

Optional login when you need it

Off by default for a laptop. Turn on a single account, or OIDC against Keycloak, Google, Entra, or Okta with a user allowlist, when it goes on a network.

Download Beta

Latest version: 0.6.0 (release notes). Docker is the recommended way to run Alginte.

Alginte is in beta (0.x): configuration keys, the topology export format, and APIs may change between releases; breaking changes are called out in release notes with migration guidance.

Recommended

Docker image

Layered image on Docker Hub — small upgrade pulls, runs anywhere.

docker pull alginte/alginte:latest

A reachable Kafka broker is required at startup — point SPRING_KAFKA_BOOTSTRAPSERVERS at your cluster.

Coming soon

Fat JAR

Self-contained JAR (~186 MB) — frontend + backend in one. The direct JAR download arrives in a future release; the Docker image is the supported channel today.

java -jar alginte.jar

No Kafka cluster handy? The playground starts Alginte with a bundled broker and sample data — one command, nothing else needed.

Full install guide and configuration reference in the documentation.

Get notified of new releases

One email per release — no marketing, unsubscribe any time. Also the first to hear when the Professional edition launches. How we handle your address: privacy.