Enterprise workflow automation platform

Workflow automation,
built to enterprise grade

From the screens people use (SSemForm) to files, workflows and AI agents — all on one platform. Install it on your own infrastructure; saga compensates failures automatically.

workflow — live execution
Webhook Trigger in AI Agent LLM · tool calls Transform Data transform Merge Merge · store Running distributed — in parallel Saga pending · 0
88
Microservices
7+
AI providers
21
Core service modules
1-Click
n8n · Make import
WHY SSEMWORKS

Automate the repetitive work,
and own the platform yourself

SSemWorks begins where SaaS automation tools stop — air-gapped networks, large-scale processing, transactional consistency.

Where existing tools fall short

  • SaaS lock-in — your data leaves the building, and air-gapped or internal networks rule the tool out entirely.
  • Single-process execution — everything runs in one process, limiting horizontal scaling and fault isolation.
  • No failure recovery — when a step fails, the work already done stays behind and has to be cleaned up by hand.
  • Governance gap — no company-wide control over AI tool usage, files and permissions.

How SSemWorks solves it

  • Self-hosted — install it on-premises or in an air-gapped network; the data and the platform stay yours.
  • Distributed execution on Kafka — every microservice runs on the message bus, so each layer scales independently.
  • Automatic Saga compensation — on failure, completed steps roll back in reverse order to keep data consistent.
  • Enterprise security — a single gateway, API-key authorization, encrypted credentials and role-based access.
CAPABILITIES

What the platform gives you out of the box

From user-facing screens to the execution core, AI governance and large files — it is all in one platform, with nothing to assemble yourself.

Distributed execution core

Coordination, planning and execution run as separate layers on the Kafka message bus. A worker failure cannot stop the engine, and each layer scales horizontally.

Saga distributed transactions

When a microservice fails, the steps already executed are compensated automatically in reverse order. Progress is tracked on a dedicated console, and you can step in by hand.

AI agents & MCP gateway

Swap models freely across multiple LLM providers, and run agents that use tools. The MCP gateway federates many tools behind a single address and authorises down to the individual tool with an API key.

Stream server

A dedicated file server takes the bytes while only a reference flows through the workflow. Load is separated in two places — on upload and between tasks — so files of hundreds of megabytes never weigh on your page servers or the engine.

n8n · Make import

Paste an existing n8n workflow JSON or a Make blueprint and it comes straight in. Service contracts, trigger semantics and expressions are compatible, so your team keeps its assets and its know-how.

Management studio

Design with a drag-and-drop visual editor and monitor executions in real time. Folders, permissions, credentials and insights are managed on one screen.

SSemForm user screens

A workflow begins and ends at a screen a person looks at. Build the forms and PDF templates yourself; signatures are bound to the submission and sealed as evidence.

Where people step in

Where a human decision is needed — approve, reject, request missing documents — the workflow pauses and waits. Once submitted, only the values you named flow onward.

Microservice registry

Developers publish the microservices they build as packages, and each site's administrator picks and installs them from a settings screen. Nobody touches the server, and what is installed is usable on the canvas at once.

Enterprise sign-in and audit

SSO over OIDC and SAML, LDAP/AD integration, two-factor authentication and role-based permissions. Even failed sign-ins are recorded in the audit trail, and secrets never are.

ARCHITECTURE

The architecture at a glance

All inbound traffic through a single gateway, all execution on a distributed engine over the message bus — a simple principle holds the whole thing up.

ENTRY Studio Visual editor · console Triggers Webhook · schedule · chat GATE API Gateway Single entry · access control CORE Workflow Engine Coordinate · plan · persist · Saga Message Bus — Kafka Worker Service exec Worker Service exec Worker Service exec SERVICES Data & Files PostgreSQL · file stream AI & MCP Multi-LLM · tool gateway

Every external call must pass the API gateway before it can reach the engine. The engine plans the run and hands microservices to workers over the message bus, while state is always persisted in the data layer — restart any component and the record of the run survives.

HOW IT WORKS

From design to compensation, in four steps

From building a workflow to running and observing it — the whole cycle closes inside the platform.

1

Design

Drag microservices together in the visual editor to compose a workflow. Existing n8n JSON and Make blueprints come straight in through Import.

2

Trigger

Webhooks, schedules, messengers, SSemForm forms and more all pass the gateway and start a workflow.

3

Distributed run

Each microservice runs on a worker, in parallel and distributed, over the message bus. Large payloads bypass the bus automatically.

4

Observe & compensate

Track the whole run in real time; when a failure is detected, Saga rolls completed steps back in reverse order.

FOUNDATION

Built on proven engineering

Built from standard technology your team already knows how to operate. No new stack to learn — your existing Java and container skills are enough to adopt it.

Java 21 · OSGi
Apache Kafka
Apache Karaf · Camel
PostgreSQL
Docker containers
React admin console

Considering SSemWorks for your organization?

We will walk you through the right setup and rollout for your environment.

Get in touch