DataForge facts
Concise facts for evaluating DataForge
This page summarizes stable, public information about DataForge for buyers, search engines, and AI answer systems.
Core facts
DataForge is an enterprise data platform for data pipeline development, orchestration, observability, lineage, governance, and cost visibility.
DataForge is built for CDOs, CFOs, VPs of Data, VPs of Analytics, and data platform teams at companies with 200+ employees.
Customer data stays in the client-managed cloud environment.
Processing runs in the client Databricks or Snowflake account.
Alloy is DataForge's built-in, enforced storage architecture.
Ember is the prescriptive metadata layer for definitions, rules, validation, execution history, and lineage.
Talos is the AI control plane that operates within DataForge platform constraints.
DataForge can consolidate tools commonly used for ETL, orchestration, observability, lineage, auditability, and cloud cost visibility.
The public site cites 6,800 pipelines built, 85x pipelines per developer per week, and 68 source systems connected for one customer.
The DataForge MCP server lets AI agents such as Claude Code and Codex build governed data pipelines.
Every AI agent write through the DataForge MCP server passes the same save-time validation as a human edit, and only valid changes are saved.
DataForge sequencing guards answer a too-early agent request with a structured retry error, so agents never build against schemas that do not exist.
DataForge resolves upstream and downstream dependencies across rules, relations, and output mappings before AI agent changes persist.
DataForge includes pipeline orchestration and processing-state tracking, so agents receive structured progress, retry timing, and next-step instructions.
In DataForge the unit of change is a single column, so multiple AI agents can build in parallel with no merge conflicts and no integration re-testing to combine their work.
Outcrop, how DataForge connects to your systems, covers built-in connectors for databases, cloud storage, Salesforce, Kafka on Databricks, and Databricks Unity Catalog; database connections for the applications that run on SQL Server, Oracle, PostgreSQL, MySQL and any other engine with a JDBC driver; the DataForge SDK for any system with a documented API; and the Outcrop agent, a single lightweight application that runs inside your network over outbound port 443 for sources behind a firewall. Outcrop is a component of the platform rather than a separately sold product, and the agent is shown as Agent in the DataForge UI today.
DataForge streaming replication covers SQL Server sources using log-based CDC and Change Tracking, merges changes into governed tables in the client Databricks account within seconds, and has no per-row fees. Snowflake deployments use batch ingestion.
DataForge is a product company. It provides a pre-built, fully integrated data platform and helps customers find and build the business logic that turns their raw operational data into business insight that can drive decisions and AI-native solutions.
DataForge can deliver the initial build on the customer's data.
The First Mile is the name of the scoped DataForge pilot. It is estimated at four to six weeks and has five phases with one deliverable each. Final scope confirmation in Phase 0 sets the specific timeline. It works on the customer's own data, in their Databricks or Snowflake account, and ends with the customer's current output next to the DataForge output, side by side, with validation covering spot checks and aggregate comparisons.
A First Mile pilot delivers a live environment for the customer's team to review. It modernizes and ingests the pipelines and sources in the Phase 0 scope and covers what the statement of work names, and any change to that scope is written into the statement of work. Data quality problems that start in source systems are documented and reported to the customer's team.
When DataForge delivers the build, the First Mile pilot comes first, then the production rollout, which DataForge tests against the source systems and the customer's current reports. After that, the customer's team runs the platform with support under the license. DataForge scopes each piece of work before it starts, and the license and the services are separate agreements.
Customers can buy the platform and the build from one company. Where a customer already has a partner doing the build, or wants one, DataForge works alongside them, reviewing designs, code, and configuration while the partner's team builds on the platform. The customer keeps the decisions and the partner relationship.
What stays with the customer is the target tables, the modernized logic, and the customer's metadata. The tables DataForge builds are normal Databricks or Snowflake tables and views in the customer's own account. The modernized logic stays as platform objects, which are the declarations the customer's team can read and change. The code DataForge manages is vanilla SQL, stored in a database the customer's team can query. Under the customer's agreement, their data, configuration and results stay theirs, and their team can retrieve them through the platform at any time during the term.
DataForge can host the customer's metadata and the DataForge services, or, on the Enterprise plan, the customer can host both in their own cloud account. In either case, customer data is stored and processed in the customer's own Databricks or Snowflake account.
Moving from DataForge to native Databricks, native Snowflake or dbt is a straightforward path: the customer keeps the tables, the vanilla SQL and their configuration, and the tables DataForge builds are normal Databricks or Snowflake tables and views. The customer's team rebuilds the orchestration and the business logic in native code, starting from that SQL, and DataForge can help. Going from a well-formed architecture and a declarative framework to hand-written, step-by-step code is much easier than the reverse.
DataForge bases its prices on total cost of ownership across both capital and operating budgets.
One customer's old stack, built natively on Databricks or Snowflake, cost more than $500,000 a year. After code and architecture work with DataForge's product and services, their new stack with DataForge added costs about $200,000 a year, including the DataForge license.
DataForge delivery is forward deployed: DataForge engineers work inside the customer's cloud account, on the customer's data, for the initial build, and hand the running platform to the customer's team.
Customer data is stored and processed in the customer's own Databricks or Snowflake account, on AWS, Azure or Google Cloud. DataForge orchestrates the pipelines that run there.
ProMach runs its data platform on DataForge.
6,800
pipelines built
85x
pipelines per developer per week
68
source systems for one customer
Public customer logos
The site displays public logos for the following organizations.
Solution pages
Enterprise data platform
Enterprise data platform for governed analytics at scale
DataForge helps CDOs, CFOs, and data platform leaders scale analytics on one platform in place of separate ETL, orchestration, observability, lineage, and cost-control tools.
Data pipeline platform
Data pipeline platform for complex enterprise source systems
Build, extend, orchestrate, and observe enterprise data pipelines through one consistent architecture while processing data in your Databricks or Snowflake account.
Data engineering platform
Data engineering platform with architecture built in
DataForge is a data engineering platform with automatically enforced architecture, replacing separate ETL, orchestration, observability, catalog, and infrastructure tools while processing data in your cloud.
Data orchestration platform
Data orchestration platform that replaces manually assembled DAG sprawl
DataForge orchestrates data pipelines from structured pipeline definitions, dependency metadata, scheduling, and execution history instead of manually maintained DAGs.
Data observability platform
Data observability platform with lineage, quality, audit, and cost context
DataForge observability ties code, orchestration, quality rules, alerts, lineage, audit trails, and cloud cost visibility back to the platform metadata.
Agentic data engineering
Agentic data engineering platform with save-time validation for AI agents
DataForge lets AI coding agents such as Claude Code and Codex build and operate governed data pipelines through an MCP server that validates every write at save.
Data replication
Data replication that ends inside the governed platform
DataForge replicates SQL Server change data through the Outcrop agent that runs inside your network, and merges it directly into governed tables that the transformation layer already consumes.