An AI coding agent for SAP development

Senior SAP developer.
At every desk.

One AI coding agent for the whole SAP stack. It works on your own SAP system, and you approve what it changes.

  • ABAP
  • RAP & CDS
  • OData services
  • Fiori apps
  • Side-by-side apps (CAP)
  • Clean core
  • S/4HANA upgrades

Works on ECC (EHP7 and later) and S/4HANA, on-premise and private cloud.

Bring CSPeach to your team

Just trying it yourself? Start free.

Example from a sample spec

You›/abap-app @customer-credit-review.docx
CSPeachStep 1 of 3 · Finding the gaps in the specI found 4 questions the spec doesn’t answer. 1 must be answered before anyone builds.
CSPeachmust answer When a customer is rejected, should SAP stop new orders, or only show a “credit hold” note?
You›Only show the note. Sales decides.
CSPeachSaved to the spec. Next:
should answer Can a manager overrule a reviewer?
You›No. Only Sarah Davis and Mike Johnson decide.
CSPeachSaved. 2 left, both can wait. Ready for step 2: choosing how to build it.
CSPeach reads the spec first and asks what a senior consultant would ask. You answer, and it builds on your answers.
The write road A write starts because you asked. For multi-object work CSPeach lays out a plan you approve, then it snapshots, writes, runs a syntax check, activates, and confirms the object really went active. If a check fails, CSPeach fixes the object and checks again, up to three tries. When it still cannot pass, or the next step is one only a person can do, it stops and says what it needs; you do that step and it carries on. The road sits in the session’s transport, or in the request that already holds the object, and stays on the record. You ask A plan you approve For multi-object work The session’s transport Or the request that already holds the object. On the record with /export audit. Snapshot Before the write Write Syntax check On every write Activate Activation check Reports if it didn’t Passes Broken code is never activated. Fails Fails CSPeach fixes it and checks again Up to three tries Tries again It stops and asks you Says what it needs Still failing, or it needs you You do the step, it carries on. The snapshot from before the write is there if you want to go back.
Every write takes this road. If a check fails, CSPeach fixes it and checks again, and asks you when it needs you.

Three kinds of work. One agent.

Start with the work your team has in front of it. Each door opens on the page that shows it.

Proved in a live team run on a real S/4HANA system

  • Two people worked one project in one shared folder.
  • Fixes were written to SAP and proven by SAP's ATC check.
  • The terminal status, the status page and the workbook show the same counts, and the CSV export carries every object row behind them.
  • Every write was approved on a card that names its transport.

Spec gaps, design, estimates, CCA, upgrade scans, the boilerplate — the senior work that eats your team's week.

Most AI tools end at suggestion. CSPeach is a teammate — it asks the right questions, runs the unglamorous checks, drafts the boilerplate, and hands the judgment calls back to the human who knows the business.

Skills that chain.

Two kinds of work, two chains. A new build starts from the spec you already have — a PDF or Word doc — and runs through the questions it doesn't answer, the design, the plan, then the build: a RAP stack and its Fiori app, or a CAP app beside the core. An upgrade of your custom code is its own road: custom-code analysis, the readiness scan, the fixes, then the sign-off report.

Skills that chain — two chains, not one Two separate chains, each in its own territory, with no arrow between them. The first is a new build from a spec: the PDF or Word doc you already have, then spec-gap for the questions the spec does not answer, then design for the object list and the build sequence, then plan for the steps, approved once with their object list, then build — the objects and the app on top of them. The second is an upgrade of your custom code: custom-code analysis reads the system inventory and proposes the scope, the upgrade scan raises ATC findings against your target release, the fix works one object at a time with each change approved, and verify runs the same scan again against the baseline. Inside each chain a phase's output is the next phase's input. Between the two chains nothing carries over — they are different projects. TWO CHAINS, NOT ONE Each phase's output is the next phase's input. A NEW BUILD · FROM A SPEC AN UPGRADE · YOUR CUSTOM CODE Your spec The document you already have PDF OR WORD DOC spec-gap The questions the spec does not answer /abap-spec-gap design The object list and the build sequence /abap-design + /abap-estimate plan Steps, one approval for the whole list /abap-plan build The objects, and the app on top of them /abap-rap /abap-fiori-build inventory custom-code analysis Reads the inventory, proposes the scope /abap-cca upgrade-scan scan ATC findings against your target release /abap-upgrade-scan upgrade-progress fix One object at a time, each change approved /abap-upgrade-fix upgrade-report verify The same scan again, against the baseline /abap-upgrade-verify Your project context never gets re-explained.
A build from a spec and an upgrade of your custom code are two separate chains — nothing carries from one to the other.

SAP writes you can stand behind.

In your development system, through the same road every time: a plan you approve, a snapshot, checks, and the session's transport.

  • Read-only by default.
  • Your own SAP system runs the syntax check and the activation.
  • Your ABAP goes straight from your laptop to SAP.
CSPeachPlan: change 3 objects in transport DEVK900123. Approve?
You›Yes.
CSPeachSnapshot taken. Writing ZCL_CREDIT_REVIEW…
Syntax check on your system: passed. Activated. Checked it really went active.
Where your data goes ABAP work goes from your laptop straight to your SAP system over ADT, with nothing in between. In every mode the CSPeach server handles sign-in and serves the signed skills and routing instructions. In managed mode, the default, prompts also go through it to the AI model. Your own model key, with prompts going straight to the model, is a future option and not offered today. The CSPeach server never reaches your SAP system. Our side Your side Sign-in and skills, every mode Prompts, in managed mode Prompts Your own model key: a future option, not offered today. Our server never reaches your SAP Your ABAP goes straight to SAP Over ADT, in every mode. Nothing in between. CSPeach server Sign-in, signed skills, routing AI model Your laptop Runs the cspeach CLI Your SAP system Runs the checks and activation
ABAP work goes straight from your laptop to your SAP system. Our server handles sign-in and skills in every mode, and carries prompts in managed mode.

More on write modes, and on how CSPeach handles your code, credentials and prompts.

35+ skills across the SAP lifecycle.

Pick the skill that matches the work. Book a demo to see them at work.

/abap-spec-gap

Find what's missing.

Reads a spec — a PDF or Word doc — and finds the questions to answer before anyone touches code.

/abap-design

Decide before coding.

The object list and the order to build it in.

/abap-estimate

Hours with risk.

Ranges per component, and the risks behind them.

/abap-clean-core

Level A to D, with evidence.

Checks released-API state and ATC on your system. Read-only.

/abap-rap

A full RAP stack, activated.

Table to service binding, created in order after a plan you approve.

/abap-cca

Scope the whole estate.

Reads the system inventory and proposes the scope and the retire candidates. Then each object is decided: keep, fix, retire or rebuild.

/abap-upgrade-*

Scan, fix, verify.

A baseline scan, guided fixes with rollback, and a sign-off report.

/abap-incident

Dump to deployment.

Short dump in, fix and verification out.

Clean core, RAP and CDS: see what CSPeach does.

A clean-core project, in eight steps.

Clean core is two jobs: new work goes in clean, and what you already have gets counted, classified, fixed and proven.

01Scope 02Split 03Measure 04Decide 05Fix 06Prove 07Report 08Keep

How a clean-core project runs

See it run on a
real system.

Book a demo to watch real sessions on a real system.

An S/4HANA upgrade, run as one team project Five stages left to right. Scope counts every custom object from your system's own inventory with cspeach team discover. Decide has each person record keep, fix, retire or rebuild for their objects with cspeach team decide. Fix gives each developer a lane, where every fix is a claim, with /abap-upgrade-fix. Prove measures each claim with a check run from cspeach team refresh: fix proven, or fix failed the check. Keep: after go-live, a later check run shows any fix that slipped back. All five stages are built. Under them, management reads the status page, the workbook and the CSV export, which show the same counts. TEAM PROJECT · TYPE: UPGRADE An S/4HANA upgrade, run as one team project. A fix counts when a check proves it. BUILT Scope Every custom object, counted by your system cspeach team discover BUILT Decide Keep, fix, retire or rebuild, object by object cspeach team decide BUILT Fix Each developer works a lane; every fix is a claim /abap-upgrade-fix BUILT Prove A check run measures it: fix proven, or fix failed the check cspeach team refresh PROVEN ON A LIVE S/4HANA SYSTEM BUILT Keep After go-live, a later check run shows any fix that slipped back cspeach team refresh WHAT MANAGEMENT SEES, ANY DAY The status page, the workbook and the CSV. The same counts. nothing to fill in
An S/4HANA upgrade run as one team project — scope, decide, fix, prove, keep. Every developer's laptop writes small status files into your own shared folder, and management opens one read-only page any day. Team projects are new; ask for them in the walkthrough.

We also keep an honest list of what's not ready yet.

Built to keep up.

AI gets better every few months. SAP development should get better with it. We keep up, so your team doesn't have to — with a person in the loop at every step.

Where CSPeach is going

Give your SAP team a senior ABAP developer on tap.

  • Works on your SAP system.
  • Your ABAP goes straight from your laptop to SAP.
  • By default, you approve every write.

Free demo and trial for SAP teams at companies in the US.

Built by Cremencing Solutions — SAP and ERP development since 2013 — and tested against a live S/4HANA system, not a slide deck.

A real walkthrough and a free trial — not a sales call. Reply within 1 business day. No spam, unsubscribe anytime.