Matillion

Overview

Matillion is a low-code ELT and transformation platform, and it is the most transformation-centric of the mainstream ingestion tools. Where others move data and leave modelling to you, Matillion's core has always been building pipelines visually and pushing the work down into your warehouse.

Pushdown is the architectural point. Matillion generates platform-specific SQL and executes it inside your data platform, so data never leaves it, and on Snowflake it can push Python down through Snowpark too.

Pipelines come in three kinds: orchestration for sequencing and loading, transformation for reshaping tables already in the warehouse, and test pipelines that run another pipeline and check its results.

Unusually for a visual tool, Git integration is native in the designer, so low-code pipelines stay version controlled and reviewable rather than trapped in an interface.

Know before you search that the product is mid-rebrand. Matillion's pages now redirect to maia.ai, and the Data Productivity Cloud has become Maia Foundation, so documentation links written before 2026 will not resolve.

Maia is the wider positioning: agentic AI described as virtual data engineers that build and maintain pipelines, with Mission Control for governance and a context engine for standards.

Leadership changed with it. Girish Pancha, formerly of Informatica and StreamSets, became chief executive in August 2026, with founder Matthew Scullion moving to the board.

Connector counts vary by page between 130 and 150, with custom connectors built through a no-code wizard against any REST API, and flex connectors built to order by Matillion.

On pricing it publishes consumption precisely and price not at all. Credit rates per task hour are in the legal terms, but no monetary value of a credit appears anywhere, so every tier is effectively quote-only.

It is fully commercial, with no open-source, open-core or source-available component of any kind.

Key features and capabilities

The same headings are used for every elt & ingestion entry, so two tools can be read side by side.

Connectors and coverage

• Counts vary by page, 150 on the Maia site against 130 in the architecture documentation
• Coverage spans databases, ERPs including SAP, CRMs, APIs, files and documents
• Custom connectors are built in a no-code wizard against any REST API, without scripting
• Flex connectors are built to a customer's request and delivered in days
• No certified or community support tiers exist, since every connector is vendor-built

How pipelines are built

• A low-code visual designer is the primary surface, assembling configured components
• Three pipeline types, orchestration, transformation, and test pipelines that verify another pipeline's results
• Native Git integration in the designer keeps visual pipelines version controlled
• Scheduling, a REST API for execution and status, and Databricks notebook orchestration are built in
• Transformation is fully included, which is the clearest contrast with the pure movement tools

Loading and sync behaviour

• Batch loading from multiple sources, plus change data capture as a first-class streaming product
• Change capture is billed in essential and premium tiers, at one and two credits a task hour
• Reverse ETL pushes enriched data back out to operational systems
• Consumption is measured in two-second increments and only while a pipeline runs; validation is free
• Deduplication and slowly changing dimensions are not published features; you build them in a pipeline

Destinations

• Snowflake, Databricks and AWS are the platforms named on first-party pages
• Snowflake is the deepest integration, with Python and Bash pushdown and Cortex ML
• Databricks adds Unity Catalog integration with automatic metadata tagging
• Redshift, BigQuery and Azure Synapse are not named as current targets, so confirm before assuming
• Because execution is pushdown, the destination is also the compute engine

Security, monitoring and lineage

• SOC 1 Type II, SOC 2 Type II, SOC 3 and ISO 27001:2022, with the audit reports published as documents
• GDPR and the UK Data Protection Act, with PCI DSS, CCPA and CSA STAR shown as logos without detail
• TLS 1.2 and 1.3 in transit, AES-256 at rest by default
• Multi-factor authentication, role-based access control, single sign-on and audit logging
• Lineage is supported, with auto-documentation and Unity Catalog metadata tagging; HIPAA is not published

Extending it yourself

• Custom connectors through a no-code wizard, and an agentic path for building them with Maia
• A REST API for pipeline execution and status, plus an API gateway component
• dbt Core runs inside Matillion's own pipeline infrastructure
• Pushdown Python through Snowpark and Bash pushdown for custom logic
• No published SDK equivalent to its rivals' connector kits

Where it runs and what it costs

• Full software as a service, with Matillion hosting the runner and calling out to your cloud
• Hybrid deployment runs your own runner inside your private cloud, on AWS or Azure
• Any number of hybrid runners across clouds, with private link connectivity
• Google Cloud is not named as a hybrid target, and no region lists are published
• Billed in credits, consumed by task hours and by additional developer users

Pricing

Price on requestQuote only; credit rates published, price is not

Three packages, Developer, Teams and Scale, all consumption-based on credits, with Teams and Scale routing to sales. Consumption is published exactly in the legal terms: one credit per task hour for orchestration, three for transformation, one or two for change capture, and 70 credits a month per extra developer. Consumption is metered in two-second increments while pipelines run. What is nowhere published is the monetary value of a credit, which the terms say sits on the order form, so you can size a workload precisely but not cost it. Trial length is not published.

Vendor pricing page →

Demos and videos

• See Maia in action, meet your virtual data engineers Product demo · matillion.com
• Building orchestration pipelines with Maia Tutorial · matillion.com

About Matillion

Matillion was founded in 2011 near Manchester by Matthew Scullion and Ed Thompson and now runs dual headquarters in Denver and Manchester. It is private and venture-backed through a Series E in September 2022 at a $1.5bn valuation, with Lightspeed, Battery, Sapphire and Scale leading earlier rounds and Snowflake, Databricks and Citi investing strategically; round amounts are not published. Leadership was reset around AI in 2026, with Girish Pancha appointed chief executive in August and Sanjay Krishnamurthi as chief development officer, both formerly of Informatica. Its own customer figures conflict, from a thousand data teams to thousands of enterprises.

Founded 2011 · Denver and Manchester · matillion.com