DRAVNby Poro-IT
PricingGet Started

// API Integration Platform

Data flows.
Drawn through.
Always.

DRAVN is a visual data integration platform with built-in AutoML. 28+ connectors, Spark execution, CUDA-accelerated model training, full data lineage, and GDPR compliance. Self-hosted or managed SaaS.

Get StartedExplore features →
// Deployment tiers
// tier 01
Lite
Single server, any OS. SQLite, PostgreSQL, or MySQL. Zero infrastructure overhead.
SQLitePostgreSQLMySQLWindows · Linux · macOS
// tier 02
Standard
Full feature set with PostgreSQL + Redis. Deploy on bare metal, GCP Cloud Run, or AWS ECS Fargate.
PostgreSQLRedisGCP · AWSFinland · Stockholm
// tier 03
Enterprise
Adds Kafka streaming, ksqlDB routing, CDC, and Spark execution. GKE or EKS Kubernetes.
KafkaksqlDBSparkGKE · EKS
// tier 04
DRAVN CloudCOMING SOON
Fully managed SaaS. No infrastructure. Subscription-based — connect and go.
ManagedSubscriptionZero ops
28+Connectors
4Execution Modes
9ML Algorithms
27Feature Areas
6Deploy Options
CUDAGPU Acceleration

Core Features

Everything from pipeline building to GDPR compliance — built in, not bolted on.

26 FEATURE AREAS
01Standard

Visual Pipeline Builder

Drag-and-drop canvas with DAG execution, parallel branch support, and real-time edge connections.

  • 50-step undo/redo, Ctrl+D duplication
  • Export/import pipelines as JSON
  • Error routing to dead-letter targets
  • Schema propagation across nodes
DAG EngineParallel ExecError Routing
02Standard

Data Catalog & Lineage

27+ discoverable assets with quality scoring. Full directed lineage graph with upstream/downstream traversal.

  • Classification: public, internal, confidential, restricted
  • PII, financial, raw, curated tags
  • SVG lineage graph with edge types
SearchTagsLineage Graph
03Standard

PII Detection & Masking

Automatic PII detection in pipeline preview data with configurable masking before data moves.

  • Pattern matching: email, phone, SSN, credit card
  • Finnish SSN regex detection
  • Preview warnings — user decides on masking
Auto-detectMaskingPreview
04Standard

Integration Documentation

Auto-generated A4 documents per pipeline with inline SVG data flow diagrams.

  • 8 sections including lineage and run history
  • Bulk docs for all pipelines
  • Browser-native Print/Save as PDF
Auto-generatedSVG DiagramsPDF
05Standard

PDF Processing

Read PDFs as a data source and write structured data to professional PDF reports.

  • Source: table, text, lines, metadata extraction
  • Target: table, report, cards, custom templates
  • {{placeholder}} template mode
ExtractionTemplatesCustom
06Standard

Monitor & Run History

Per-node execution metrics, live auto-refresh, expandable run detail with row statistics.

  • Status badges: success, failed, running, cancelled
  • Per-node DAG results
  • Execution mode tracking per run
Live MetricsRow StatsAudit Trail
07Standard

Multi-Tenant Architecture

Organisation-level data isolation with department access control and plan-based limits.

  • Plan limits: Free, Standard, Enterprise
  • Custom domain per tenant (SaaS)
  • RBAC: org_admin, dept_admin, user
IsolationPlan LimitsRBAC
08Standard

Command Palette & UX

Keyboard-first platform with Cmd+K navigation, bulk operations, and full ARIA accessibility.

  • Fuzzy search across 18 nav items
  • Shift+Click range select for bulk delete
  • Dark mode throughout, 17 dashboard pages
Cmd+KKeyboard-firstARIA
09Enterprise

AutoML & Inference

Train multiple models on your data, compare metrics, deploy the best as a pipeline transform. CUDA GPU acceleration for enterprise.

  • 9 algorithms: XGBoost, LightGBM, Random Forest, Neural Net, and more
  • Compare models by accuracy, F1, R² — pick the best automatically
  • GPU training: XGBoost CUDA, LightGBM GPU, PyTorch with CUDA
  • Deploy trained models as inline pipeline transform nodes
AutoMLCUDAPyTorchscikit-learn
10Enterprise

Conditional Router & CDC

Conditional routing with ksqlDB. Debezium-powered change data capture.

  • Filter-based row routing to different targets
  • ksqlDB streaming SQL
  • Row-level CDC from any RDBMS
ksqlDBDebeziumKafka
// Connectors

28+ Sources & Targets

Connect to databases, streaming platforms, cloud storage, APIs, files, and more.

// Data Sources
Databases
PostgreSQLMySQLOracleSQL ServerSQLiteJDBCApache Impala
Streaming & Events
Apache KafkaksqlDBCDC / Debezium
APIs & Web
REST APIAPI SourceWeb ScraperSOAP / X-Road
Files & Cloud
Local filesPDFGCSS3Azure BlobHDFSSFTP / FTP
Scripting
Python scripts
// Data Targets
Databases
PostgreSQLMySQLSQLiteJDBCImpala + HDFS
Streaming
Apache Kafka
APIs
REST API
Files & Cloud
Local filesPDFGCSS3Azure BlobSFTP / FTP
Scripting & Other
Python scriptsEmail
// Execution

Four Execution Backends

Run pipelines where the data lives. DRAVN selects the optimal backend automatically.

01
In-Process
Node.js DAG runner with streaming chunks, batch writes, and backpressure. 500-row batches = 100–500× fewer DB round-trips.
Default · All tiers
02
Database (ELT)
Pushes SQL directly to the database. Zero data movement. Fastest path for same-database transforms.
ELT · Zero movement
03
Apache Spark
Generates PySpark scripts and submits via spark-submit for massive datasets.
Enterprise · Massive scale
04
Auto
Intelligently selects the optimal backend based on pipeline topology and dataset size.
Recommended
// Security

Security by default

Every layer built with security as a requirement, not an afterthought.

JWT + Refresh Tokens
1-hour access tokens, 7-day refresh with rotation and DB-backed revocation.
AES-256-GCM Credentials
All stored credentials encrypted at rest with unique salt per operation.
bcrypt + Account Lockout
12+ rounds. Lockout after 5 failed attempts with 15-minute cooldown.
Rate Limiting
Redis + in-memory fallback. Fail-closed on auth endpoints.
RBAC
org_admin, dept_admin, user roles with org and department isolation.
CSRF + CSP + HSTS
Origin validation, Content Security Policy, full security header suite.
Python Sandboxing
Process isolation, import whitelist, resource limits, temp directory confinement.
SQL Injection Prevention
Identifier validation. Path traversal protection with directory allowlist.
Credential Redaction
Passwords shown as dots in API responses. Error logs sanitised.
// Deployment

Runs where you need it

Fully self-hosted. Your data never leaves your infrastructure. Nordic regions by default for EU data residency.

Self-hosted
Single Server
Windows, Linux, or macOS. No infrastructure expertise required.
· SQLite · PostgreSQL · MySQL· Kafka + ksqlDB (Enterprise)
Cloud
GCP · AWS
Serverless containers with Terraform. Auto-scale with load.
· GCP Cloud Run (Finland)· AWS ECS Fargate (Stockholm)· Terraform configs included
Enterprise
Kubernetes
GKE, EKS, or any Kubernetes cluster with Helm.
· Helm chart with configurable values· VPC · private networking · IAM
01
Self-hosted
Single server
02
Cloud Run
GCP · Finland
03
ECS Fargate
AWS · Stockholm
04
GKE
Enterprise GCP
05
EKS
Enterprise AWS
06
Generic K8s
Any cloud
// GDPR & Privacy

Compliance built in

Self-hosted by design. Your data stays on your infrastructure. GDPR features built into the platform.

Data never leaves your infrastructure
No telemetry, no call-home, no third-party data processors by default.
EU/EEA data residency by default
Cloud deployments default to Finland and Stockholm.
PII detection and masking
Automatic detection before data is written to any target.
Encrypted credentials at rest
AES-256-GCM with unique salt per operation.
// Data Subject Rights

Built-in rights management

Right to Access
Data export API — full JSON download.
Right to Erasure
Account deletion with cascading deletes.
Right to Rectification
Profile editing endpoint.
Data Retention Policy
90/180 day automated cleanup. Configurable.
Consent & Audit Trail
Cookie consent logging. DPA template included.
Breach Notification
72-hour incident response procedure.
// Pricing

Self-hosted or fully managed

Run DRAVN on your own infrastructure, or let us handle everything with DRAVN Cloud.

Lite
Single server
  • —Visual Pipeline Builder
  • —15+ source connectors
  • —13+ target connectors
  • —SQL, Python, Column Mapping transforms
  • —Data Catalog and Lineage
  • —PII Detection and Masking
  • —PDF source and target
  • —Web Scraper
  • —SQLite · PostgreSQL · MySQL
Standard
Everything in Lite, plus
  • —Multi-tenant — org + dept isolation
  • —Integration Documentation
  • —GCP Cloud Run (Finland)
  • —AWS ECS Fargate (Stockholm)
  • —PostgreSQL + Redis
  • —GDPR compliance features
  • —Full security suite
Enterprise
Everything in Standard, plus
  • —AutoML — train, compare, deploy models
  • —CUDA GPU — XGBoost, LightGBM, PyTorch
  • —Decision Router — ksqlDB routing
  • —CDC — Debezium change capture
  • —Apache Spark execution
  • —Kafka Streams
  • —GKE · EKS · Kubernetes
  • —Priority support
// DRAVN CLOUD

Fully managed. Coming soon.

No infrastructure. No Terraform. Sign up, connect, and start building pipelines in minutes.

COMING SOON
// Get started

Ready to integrate
your data?

Tell us about your data infrastructure needs. Whether you want to deploy DRAVN self-hosted, use our managed SaaS, or need a custom integration project — we'll get back to you within 24 hours.

email info@poro-it.com
company Poro-IT OÜ · Tallinn, Estonia

Opens your email client. We'll reply within 24 hours.

We use cookies to analyse site traffic and improve your experience. Essential cookies are always active. Read our Privacy Policy for details.