ACBDiDATAP.AI
Data & AI sovereignty

Your data is your asset.
Not your engine vendor’s.

A vendor-neutral layer that makes your data AI-ready and keeps it, its lineage and your AI independent of any engine: one Iceberg copy in your cloud, any engine on top, governance in your own tables.

Engines govern what lives inside them. We govern what crosses between them.

Data Sovereignty →See the lineage live
YOUR PEOPLE & AIBIAnalystsAI assistantsAppsYOUR VPC · YOUR CLOUD ACCOUNTCOMPUTE · any engine, swap any timeSnowflakeDatabricksFabricRedshiftBigQuerySTORAGEApache Icebergone open copy · S3 · ADLS · GCSDATAP.AIGOVERNANCE LAYERLineagePersonal dataQualityAI guardrailsAI agents+ frameworkLegacy ETLone-off migration by DATAP.AI's AI agents & framework

From legacy to an open data & AI platform. Migrated by AI agents.

Our AI agents move Informatica, Control-M and legacy SQL to open-source dbt Core for transformation and Apache Airflow for orchestration on open Iceberg — assessed before you commit, proven before you switch, governed after.

YOUR LEGACY ESTATEYOUR NEW PLATFORMInformaticaControl-MOracleSQL ServerTeradataAzure SynapseAmazon RedshiftHadoop / HiveCONTAINERS IN YOUR VPC · FROM DAY ONEDATAP.AIMIGRATION · GOVERNANCE · AIAgentic AIDiscover · Assess · ConvertResolve · Prove · Cut overDeterministic first · AI on the long tailyour engineers approveData & AI GovernanceLineage · personal-data mapQuality tests · parity evidenceAI guardrails · audit ledgerin your cloudAirflowORCHESTRATIONopen sourceschedules · dependenciescalendars · rerunsruns dbtdbtCoreTRANSFORMmodelstests · docsopen sourceCOMPUTE · swap any engineSnowflakedbt-snowflakeDatabricksdbt-databricksMicrosoft Fabricdbt-fabricAthena · Redshiftdbt-athena · dbt-redshiftBigQuerydbt-bigquerydecoupledSTORAGEApache Icebergone open copy in your cloud · S3 · ADLS · GCSevery engine reads the same tables — no second migration
During the migration — DATAP.AI's AI agents run as containers inside your VPC and move your estate, once. We never access your network or store your data.
After go-live — the same deployment stays in your VPC, owned and operated by you. Your data never flows through it.
Data platform migration →

One console. Storage decoupled from the engines, across AWS, Azure and Google Cloud.

platform.datap.ai — your data, catalogs, engines, governance and AI in one place, inside your own accounts.

Open platform.datap.ai →Architecture

Sovereignty is the reason. Iceberg is the foundation.

AI that inherits the governancemodels you choose · where you choose · gated by the data's classificationAny engine — replaceable per workloadsame models · no data moved · add one, leave oneAthenaSnowflakeDatabricksTrino…One copy on Apache Iceberg — in storage you ownS3 · Azure Data Lake Storage · Google Cloud Storage · open catalogyour account · your region · yours before, during and afterDATAP.AI LAYERbeside the engines · nothing flows throughlineagecolumn level, across enginespersonal dataflags that follow the dataquality evidencetests born with the pipelineAI gateallow · conditions · block · escalateevidence written to your own tablesDashed: the layer reads catalogs and artefacts. Solid: the gate in front of your AI. Switch the layer off and storage, engines and evidence tables stay where they are.Built and measured on our own environment, September 2026 — three engines, two clouds, one copy.
01Open storage you own
One Iceberg copy · S3 · ADLS · GCS
02Any engine, replaceable
Same models · no data moved
03Governance that travels
Lineage · personal data · quality · your tables
04AI that inherits it
Models you choose · gated · evidenced

AI-ready. Governed data first, then the AI on top.

AI-powered pipelines
Structured tables and documents in; tested, governed Iceberg tables out. Models and tests written from the source, reviewed by your team.
Pipelines →
AI that inherits the governance
Models you choose, running where you choose — cloud or local — answering only from data whose lineage and classification you hold.
AI Sovereignty →
The governance gate
Every request checked against a versioned control catalogue before it reaches a model: allow, conditions, block or escalate, with citations.
AI Governance →
platform.datap.ai
The console that brings your data and AI services together in one place, across AWS, Azure and Google Cloud, inside your own accounts.
Open the platform →
How it fits together: Architecture

One lineage. Three engines, two clouds, three teams.

It walks through columns on its own; click any column to take over. Built on our own environment.

3
engines · 2 clouds
1
copy of the data
46
column edges crossing engines
0
unresolved columns

Sample data, our demo environment, 22 September 2026. Not a benchmark.

Built before written about.

3-minute read
Your data is your asset. Not your engine vendor’s.
Engineering note 1
Three engines, one shared Iceberg copy, one lineage
Engineering note 2
Three teams, three repositories, one lineage
Code on GitHub: estate · producer · analytics · ml
AWSAzureGoogle CloudAPRAASICMASHKMA · SFC8 APAC languages
Your account, your region, your regulator. Designed for; not a certification claim.
platform.datap.ai ↗stock.datap.ai ↗gov.datap.ai ↗
Where it runs today.

Point us at ten tables. Keep everything we return.

Column-level lineage, a personal-data map and a first generated pipeline — on your engine, in your account.

Start with ten tables