Assessments · agentic

Three assessment agents to get your SAP estate ready

Each one is an agent, not a questionnaire. A supervisor routes specialist agents across the estate read-only, a human validates the flags, and a scoring agent hands you a costed, prioritized roadmap. They share one knowledge graph and the same governance with the agents that run your support, so an assessment flows straight into the work it recommends.

01 · Migration agent

ECC to S/4HANA readiness

An agent that scores custom-code health, table complexity, configuration footprint and integration risk across the estate, then recommends greenfield, brownfield or selective data transition with a costed roadmap.

Benefit: a de-risked, smaller-scope migration.

Details →
02 · AMS / automation agent

What to automate next

An agent that mines live ticket history for recurring issues, ranks them by volume and resolution time, and proposes the candidate agents and prevention plays with the highest return.

Benefit: a measurably lower cost to serve.

See how it works →
03 · Data & AI agent

Lakehouse & AI feasibility

An agent that scores data quality, master data, lineage and governance, then rates readiness to feed Datasphere, Business Data Cloud or Databricks and to deploy AI safely.

Benefit: a trustworthy, governed data foundation.

Details →
01 · Migration readiness agent

ECC to S/4HANA, scored and costed

A supervisor routes specialist agents over the estate, a human validates the flags, and a scoring agent turns four weighted dimensions into a 0 to 100 readiness score and a phased, costed roadmap. The biggest lever: 40 to 60% of custom code is typically unused and can be decommissioned, not migrated.

30% weight

Custom-code health

Z-objects, modifications and deprecated APIs scanned via ATC and the Simplification database, scoped by real usage.

25% weight

Table complexity

Volumes, growth and archiving candidates that drive the data-migration effort.

25% weight

Integration risk

Interface count, adapter types and external connections, the main technical risk.

20% weight

Configuration footprint

Customizing scope, which sets process and testing complexity.

Why this agent, not a code scanner
One agent, every dimension

A roadmap, not a backlog

Code scanners stop at the ABAP layer. This agent orchestrates code, data, configuration and integration into one weighted, costed, board-ready roadmap.

Decommission-first

Delete, don't migrate

It proves which custom code is actually unused, so you retire it instead of paying to remediate and carry it forward.

Built on the estate it runs

Grounded in real usage

The assessment reads the same production telemetry our run-agents collect, and the agents that assess hand straight off to the agents that support you after cutover.

Trust boundary by design

Metadata only

Claude reasons over code structure and catalogs; your business data and PII never leave the tenant. Scoring runs in your environment.

After the assessment, you

Pick the path (greenfield, brownfield or selective data transition), sign off the dead-code decommission list, and fund a phased program, with the run-agents staged to execute and support it.

Benefit

A de-risked, smaller-scope migration: less code to carry, a defensible cost and timeline, and continuity into AMS once live.

Output: readiness score, decommission list with savings, remediation backlog split functional-redesign vs technical-adaptation, path recommendation, costed phased roadmap, risk register. Figures are modeled and illustrative, validated per engagement.

02 · AMS / automation readiness agent

What to automate next, ranked by return

This agent already runs in the product. It reads the live ticket history read-only, clusters recurring issues into signatures, ranks them by volume and resolution time, and proposes the candidate agent for each plus the prevention plays that stop tickets before they open, laid out as a phased rollout.

Find

Recurring signatures

Cluster the queue into the issues that actually repeat.

Rank

By volume and MTTR

Surface where automation returns the most, fastest.

Recommend

Candidate agents

The agent and prevention play to deploy for each signature.

Sequence

A phased rollout

Quick wins, then approved agents, then prevention.

After the assessment, you

Approve which candidate agents to deploy and in what order, set the autonomy and approval thresholds, and turn on the first prevention plays.

Benefit

A measurably lower cost to serve: fewer repeat tickets, faster resolution, and engineer time redirected from firefighting to higher-value work.

03 · Data & AI readiness agent

Ready for the lakehouse, ready for AI

SAP's data future is Business Data Cloud (built with Databricks) and Datasphere, with BW/4HANA maintenance ending in the same window. This agent scores whether the data is clean, governed and traceable enough to feed a lakehouse and to deploy AI safely. Governance and lineage are first-class, not a footnote.

Data fitness

Quality & master data

Completeness, duplicates and validity by domain; MDG hygiene across vendor, customer, material and finance; BW debt to retire or carry forward.

Governance & lineage · SAP-native

Traceable end to end

Lineage via the Datasphere Catalog (source to BW to views to consumption, with impact analysis), MDG, Information Steward, Data Access Controls plus S/4 authorizations, ILM retention, and the Unity Catalog hand-off into Business Data Cloud / Databricks.

AI feasibility

Use cases worth doing

Each candidate scored on value vs data-readiness, including Joule and custom-agent fit, with the PII boundary made explicit.

Scores readiness for: SAP Datasphere · SAP Business Data Cloud · Databricks · Snowflake · Microsoft Fabric · Joule & custom agents

Why this agent, not a data audit
Agentic, end to end

Profile, trace, score

Specialist agents profile data, trace lineage and score AI use cases under one supervisor with human approval, not a one-off snapshot.

Lineage you can prove

Source to consumption

Governed and traceable from the SAP source through to where AI consumes it, so you can trust what you feed a model.

A data heritage

We build it, not just report it

Backed by NuStudio's background in data and analytics, the same team designs the lakehouse and the models the assessment recommends.

Trust boundary by design

Metadata only

Claude reasons over schemas and profiling statistics; business data and PII never leave the tenant. Profiling and scoring ML run in your environment.

After the assessment, you

Fix the top master-data and lineage gaps, pick the lakehouse target and the first AI use cases to fund, and get a governed extraction plan, with the remediation and build executed for you.

Benefit

A trustworthy, governed, traceable data foundation: AI and Joule deployed safely and sooner, with the PII boundary proven, not assumed.

Trust boundary: Claude reasons over metadata only (schemas, profiling statistics, lineage, catalogs). Business data and PII never leave the tenant; profiling and scoring run in the client environment. Targets and figures are illustrative.

Start with an assessment

See what your SAP estate is ready for.

A fixed-scope, fixed-fee read of the estate, scored and costed, with a roadmap you can fund. The findings flow straight into the agents that do the work.