The Autonomous Data Operating System

From data pipelines to autonomous systems.

Replace 5+ fragmented tools with one intelligent platform. Hybridyn unifies pipelines, lakehouse, AI, governance, and serving — so your data systems build and run themselves.

Built with
PythonReactDuckDBFlinkOllama
Full stack →
5 min
To Deploy
5+
Tools Replaced
9
Architecture Planes
62
Data Connectors
9
AI Providers
593+
API Endpoints
The Problem

Data stacks are fragmented.

A pipeline tool here. A catalog there. Governance as an afterthought. AI bolted on top. Teams spend more time wiring tools together than building what matters.

5–10 disconnected tools per stack
Manual orchestration across every layer
No unified view of data health
AI locked in separate silos
Months to onboard, weeks to debug
The Shift

Hybridyn replaces the stack.

One platform. Nine purpose-built planes. Everything integrated from day one — pipelines, storage, AI, governance, serving, and monitoring. No glue code. No duct tape.

1 platform replaces 5+ tools
Autonomous pipeline execution
Platform-wide health scores and SLAs
AI native in every layer (9 providers)
Deploy in minutes, not months
Why Not Stitch Tools Together?

Traditional stack vs. Hybridyn

AspectTraditional StackHybridyn
Tools needed5–10 separate products1 unified platform
OrchestrationManual wiring + scriptsAutonomous execution
AI capabilitiesSeparate layer / noneNative across all 9 planes
Data governanceBolted on after the factBuilt in from day one
Time to deployWeeks to monthsMinutes (Docker Compose)
ObservabilityPer-tool dashboardsPlatform-wide health score
Cost5+ licenses + integrationSingle platform
MaintenanceN tools × N upgradesOne system, one upgrade
The 9-Plane Architecture

Purpose-built. Deeply integrated.

Three functional clusters — each plane handles one concern, all planes work as one system.

Core — Move Data
Connectors
62 providers across 12 categories. Wizard-based setup.
Execution
Priority queues, engine routing, resource quotas.
Lakehouse
Medallion architecture. Bronze → Silver → Gold with schema enforcement.
Intelligence — Understand Data
AI
9 providers. NL→SQL, agents, code gen. Local LLMs for privacy.
Catalog
Discover, profile, search. Automated lineage and tagging.
Serving
Data products, QueryLab, data marts. REST API publishing.
Control — Govern & Monitor
Governance
5-tier RBAC, data contracts, classification, compliance.
Observability
Health scores, SLA tracking, metrics, alerting.
Pipeline
Visual builder, templates, scheduling, expression editor.
Unified Control Plane
AI-Native

Describe it. Hybridyn builds it.

AI isn't a chatbot sitting in the corner. It's embedded in every plane — creating pipelines from descriptions, querying data in plain English, and debugging failures automatically.

"Create a sales pipeline"
AI builds nodes on canvas, configures connections, sets scheduling.
"Why did yesterday's job fail?"
Root cause analysis with fix suggestions. One-click retry.
"Show me revenue by region"
NL→SQL generates the query, executes it, returns results instantly.
Privacy-first: Run Ollama locally to keep data in your VPC. No data leaves your infra.
$ hybridyn ai status
Claude
OpenAI
Gemini
Azure
Ollama
DeepSeek
Mistral
Groq
Custom
> What AI agents can do
Create pipelines from natural language
Execute multi-step workflows (MCP)
Query data without writing SQL
Debug failures with root cause analysis
7 tools + workspace custom tools
> Provider detection: auto from env vars
> No provider? Smart stub mode (works offline)
terminal
$ git clone https://github.com/hybridyn/fpulse
$ cd fpulse
$ docker compose up -d
✓ PostgreSQL ready
✓ Redis ready
✓ MinIO ready
✓ Redpanda ready
✓ F-Pulse API ready
✓ F-Pulse UI ready
$ open http://localhost:5173
Pipeline canvas ready. Start building.
Developer Experience

Running in 5 minutes. Not 5 weeks.

Infrastructure-in-a-box. 22 pre-configured services orchestrated with a single Docker Compose command. No manual setup, no dependency hell, no configuration marathon.

Single command to full platform
22 services, pre-configured and orchestrated
Visual pipeline builder at localhost:5173
API explorer at localhost:8000/docs
Hot-reload for development
Products

Free for builders. Enterprise for teams.

F-Pulse

Free & Open Source (MIT)

Visual pipeline builder, universal connectors, SQL/Python transforms, scheduling, and monitoring. Everything you need to move data — free forever.

Best for: Individual developers, small teams, startups
Visual BuilderConnectorsSchedulingTemplatesAI Planning
Download free

D-Pulse

Enterprise Platform

The complete Data Operating System. Everything in F-Pulse plus lakehouse, AI agents, governance, catalog, serving, multi-tenancy, and 593+ API endpoints.

Best for: Platform teams, enterprises, data-intensive organizations
All F-PulseAI AgentsLakehouseGovernanceCatalogMulti-Tenant
Request demo
When to upgrade? When you need lakehouse storage, AI agents, data governance, catalog, multi-tenancy, or enterprise SSO — upgrade to D-Pulse.
See full comparison →
Who It's For

Built for every data role

Data Engineers

Replace scattered scripts with visual pipelines. Medallion ETL templates, auto-retry, built-in monitoring.

Medallion ETLTemplatesMonitoring
F-Pulse + D-Pulse

Analysts & BI

Self-serve data via QueryLab. Ask questions in plain English. Build data marts without engineering help.

QueryLabNL→SQLData Marts
D-Pulse

Governance Teams

Enforce RBAC policies automatically. Track lineage across every transformation. Classify sensitive data.

RBAC PoliciesLineageClassification
D-Pulse

Platform Engineers

Multi-tenant workspaces. Workload quotas, SSO, GitOps deployments. Infrastructure-as-code.

Multi-TenancyGitOpsSSO
D-Pulse
Technology Architecture
Compute
Python 3.11
FastAPI
React 18
TypeScript
Storage
PostgreSQL
MinIO
DuckDB
Redis
Processing
Apache Flink
Trino
Redpanda
Kestra
Security
Keycloak
OPA
Vault
RBAC

Your data deserves an operating system.

Start free with F-Pulse. Scale to D-Pulse when you're ready. No lock-in, no surprise pricing.

F-Pulse is MIT licensed. Free forever. No credit card required.