Lima connects to the systems you already run: SAP, Oracle, other databases, files, streams and applications. It reads their data in place for questions and dashboards, moves it when a system needs a copy delivered on time, and brings results back into the ERP. One catalog, one lineage record and one set of access rules cover all of it.
Six capabilities. Each one has its own page with labelled evidence.
Connects to your systems, catalogs what is there, and answers questions by reading the data in place. One small query goes to each system and the results are merged on the platform. The data stays in the system that owns it.
Open the Data Fabric →New hubMoves data between your systems, transforms it on the way, and brings results back into SAP or Oracle, from a nightly copy to a planning integration. Pipelines, orchestration, change capture and write-back.
Open the Backbone hub →One set of supply-chain terms, KPI definitions and hierarchies that every layer shares, so that a question, a dashboard and a pipeline mean the same thing by “on time and in full”.
Open the Ontology →Master data management: one golden record for each customer, material and supplier. Duplicates are matched and merged with a second person’s approval, and corrected records are written back to the system they came from. It works beside the MDM system you already own.
Open MDM →Applications built on live SAP, Oracle and WMS rows: a control tower, supply-chain apps, approval hierarchies and decision logs, with screens generated from rules.
Open SmartApps →ChatLima answers questions in plain language from the catalog, with the access policy applied and the SQL shown on request. Governed agents propose actions for people to approve, and every action is recorded.
Open ChatLima →Read it in place, or move it. One rule decides which.
The rule is short. Read in place for questions, dashboards and assistants. Move the data when a system needs a prepared, checked copy inside a time window, when a calculation is too heavy to run at query time, when a fixed snapshot is required, or when a system is being migrated or retired. Both ways use the same catalog, lineage record, quality rules and access policy, so the choice depends only on the job.
| What you want to do | Read it in place (Data Fabric) | Move it (Backbone) | Why |
|---|---|---|---|
| Ask questions, build dashboards, use ChatLima | Yes | — | The data is read where it lives and the answer comes back in seconds. |
| Feed a planning application | — | Yes | Planning engines load a prepared dataset inside a time window. They need the data delivered, checked and on time. |
| Calculate at MRP scale | — | Yes | Material requirements planning (MRP) explodes bills of material for millions of item-plant combinations. That is heavy computation, which the parallel engine is for. |
| Keep a month-end or regulatory snapshot | — | Yes | An auditor needs a fixed copy with a hash, kept as evidence. |
| Archive, offload or retire a system | Read the archive | Move it once | The Backbone moves the history. The Fabric keeps it readable afterwards. |
The full decision guide, with the economics of a copy → · What the Backbone does when the data has to move →
The same rules for a question and for a pipeline.
The platform keeps one set of rules and applies it in both directions. A question on the Data Fabric and a pipeline on the Backbone pass through the same catalog, the same access policy, the same quality rules and the same lineage record. Adding the Backbone to an estate that runs the Fabric adds no second place to govern.
Every table, view, pipeline, data product and agent in one register. The showcase estate discovered 4,284 assets in week 1.
Where each number came from, recorded column by column as questions and pipelines run. The record is current because it comes from the run itself.
Rules written in plain language and previewed on live rows. A rule that fails holds a pipeline and sends the exception to the person who owns it.
Role- and attribute-based rules at row and column level, applied before a question runs and inside every pipeline. Personal data is tagged and masked by role. Every access is recorded.
What runs on top of the data, in three plain paragraphs.
Three things run on top of the data: an assistant that answers questions, applications that run a process, and agents that propose actions. Each one reads through the Fabric or the Backbone, so the catalog, the lineage and the access policy apply to it in the same way.
ChatLima answers questions in plain language. It reads the catalog and the ontology to turn the question into a query, applies the person’s access policy, and shows the SQL on request. On the showcase estate an answer takes 4.2 seconds on average.
SmartApps are applications built on the platform’s data: a control tower with the KPI library and exception management, and supply-chain apps for approvals and decisions. They read live rows from SAP, Oracle and the WMS, and every decision is logged.
A governed agent watches the data, proposes an action such as a purchase-order draft or a corrected record, and waits for a person to approve it. Every proposal, approval and action is recorded. Its permissions are granted per tool and can be withdrawn at any time.
Read in place, extract by the vendor’s rules, write results back.
An ERP is the system that holds orders, stock, materials and finance. The platform does three things with it: reads it in place for questions, extracts from it for pipelines by the rules the vendor sets, and writes results back with a confirmation for every write. The table shows each of the three on SAP and on Oracle.
| What the platform does | SAP ECC and S/4HANA | Oracle E-Business Suite, Oracle ERP and Oracle Cloud |
|---|---|---|
| Reads in place | A question sends SAP only its filter and its columns through a released interface, with an RFC user that has read authorisations. The matching rows come back. | A question runs as a SQL query inside the Oracle database, with a read-only database user. The matching rows come back. Oracle Cloud is read through its released APIs. |
| Extracts by the vendor’s rules | Tables through released function modules, BAPIs and CDS views. Changed rows through application-level change capture. The interfaces SAP restricts in its Notes are left out. | Tables through JDBC with parallel partitions. Changed rows by last-update columns, or through LogMiner change capture where the database administrator allows it. Oracle Cloud through released APIs and extracts. |
| Writes results back and captures the confirmation | BAPI and IDoc transactions, for example the released BAPI for a planned order. Every write carries a unique key, and SAP’s confirmation is captured. | Open interface tables and public APIs for E-Business Suite and Oracle ERP, and REST APIs for Oracle Cloud. Every write carries a unique key, and Oracle’s confirmation is captured. |
ERP extraction, method by method → · Write-back and confirmation →
The four pages a review board reads first.
Security, privacy operations, reliability history and certifications, with dates.
Open →The interfaces and standards supported, each with a status and a date.
Open →Lima-managed cloud, your cloud account or your own servers, with reference configurations.
Open →High availability, disaster recovery, APIs, cost attribution and the 22-day onboarding path.
Open →The Backbone is licensed on its own as a subscription. It comes with the catalog, the lineage record, the quality rules and the access policy either way, so a team that starts with pipelines can add the Data Fabric, MDM or ChatLima later on the same setup.
SAP ECC and S/4HANA, Oracle E-Business Suite, Oracle ERP and Oracle Cloud, PeopleSoft and Dynamics 365, each through the interfaces its vendor supports. Other databases, files, streams and applications connect through the 40+ connectors or the connector SDK.
As a service on Lima-managed AWS, in your own cloud account, or on your own servers. Each installation is single-tenant, runs across three availability zones, connects to SAP over a private link, and uses your single sign-on. The Deployment page lists the reference configurations.
By the estate and the capacity you run on, as a subscription. There is no charge per row, per connector, per run or per question. Fixed-price entry offers give a trial a clear end point, such as a four-week first load with a reconciliation report.
Yes. Lima tasks run inside your Airflow DAGs, your enterprise scheduler can trigger Lima jobs, dbt runs in the same schedule, and Power BI and Tableau read certified views. Your catalog receives Lima’s lineage through OpenLineage, and your identity system provisions users through SCIM.
A thirty-minute walkthrough on your SAP or Oracle estate, then a scoped offer: a four-week Backbone first load or a 22-day control-tower pilot.
