The AI-native, agentic AI platform for SAP

Is C16 an AI-native platform for SAP?

Yes. C16 is an AI-native platform for SAP: the AI is the interface, so you work with your SAP landscape in natural language instead of menus and transaction codes. It connects to your live SAP system, ECC or S/4HANA, and produces real, system-specific results rather than generic templates. C16 is built by PEOL Technologies, an SAP Partner.

Is C16 agentic, and is that safe?

Yes. C16 is agentic: it reasons about your SAP landscape, acts, checks its own work, and self-corrects to reach a finished result. It is also governed by design: read-only by default, working strictly within your own SAP authorizations, with nothing changed in your system without your explicit approval. That pairing of agentic capability with a human approval gate is what makes it safe for enterprise SAP.

Does C16 change my SAP system? Is it safe to connect it to my live SAP?

By default, no. C16 is read-only by default and operates within your own SAP authorizations, so you decide exactly what it can access. Anything that writes back, such as deploying generated code, requires your explicit approval first. Credentials are encrypted, and PEOL Technologies, which builds C16, is ISO 27001 certified and SOC 2 Type II compliant.

Does C16 work on SAP ECC, or only S/4HANA?

Both. C16 connects to your live SAP system whether you are on ECC, mid-migration, or already on S/4HANA. Because it works on the SAP you already run, you get value today without waiting for a migration.

Do I need to migrate to S/4HANA first?

No. C16 works on your current SAP estate, including ECC, so the value is not gated on your S/4HANA migration. It actually helps de-risk and accelerate that migration when you are ready, by assessing custom code, scoring Clean Core compliance, and mapping the path.

What does C16 actually produce?

Real, system-specific deliverables generated from your live SAP data: S/4HANA readiness and Clean Core assessments, custom-code and Fit-to-Standard analysis, generated RAP and CDS code, automated tests, functional and technical documentation, and security and authorization audits. Everything is built from your actual configuration and code, not templates, and your experts review it before anything is acted on.

Does C16 train on my SAP data?

No. Your SAP data is used only to answer your questions in real time. It is never stored for model training and never shared with other customers. Each conversation starts fresh unless you explicitly save context.

How is C16 different from SAP Joule?

Joule helps users work inside SAP's screens. C16 connects to your SAP from outside, reads your actual code and data, and produces finished, deployable results under your approval. They solve different problems.

How is C16 different from GitHub Copilot or other code assistants?

Code assistants suggest code snippets in your IDE. C16 is connected to your live SAP system. It reads your actual programs, understands your data model, generates code that references your real tables and structures, validates it against SAP's compiler, and self-corrects errors. It works end to end on your real SAP, and nothing is deployed without your approval.

How long does it take to connect C16 to my SAP system?

Typically under an hour. You create a service user in SAP with specific authorizations (we provide the role template), enter the credentials in C16, and you are connected. Read-only analysis needs nothing installed and no middleware. If you later want C16 to deploy code, that adds a one-time lightweight handler that works through your own transport system.

What SAP authorizations does C16 need?

C16 uses a dedicated service user that you control. We provide a least-privilege role template covering the access C16 needs to read and, with your approval, write to your system. C16 cannot exceed what that user is allowed to do. Your SAP authorization model is the boundary.

Does C16 hallucinate? Can I trust the output?

C16 is designed to minimize hallucination by grounding every response in live SAP data. It reads your actual tables, your actual code, your actual configuration, then generates output from that evidence. For code generation, SAP's own compiler is the validator: C16 validates syntax, reads real errors, and fixes them. You review before anything is final.

What happens if C16 generates incorrect code?

Three safeguards: (1) C16 generates and validates in a controlled environment; nothing reaches your production system without your approval. (2) SAP's own compiler checks the syntax, and C16 reads those errors and corrects them automatically. (3) You review the final result before it goes live.

What is the typical use case for C16?

The most common: accelerating S/4HANA migration by assessing custom code, scoring Clean Core compliance, and generating remediation. Also used for clearing the custom development backlog, automating technical documentation, investigating process and data quality issues, and building new Fiori apps. The pattern is the same. Tasks that used to take days now take a conversation.

Who uses C16: developers, consultants, managers, or executives?

All of them. Developers use it to write and refactor ABAP, generate applications, and clear backlogs faster. Functional consultants use it to investigate process issues, assess Fit-to-Standard, and produce documentation. Managers use it to get visibility into custom code health, migration readiness, and technical debt, without waiting for a project. CXOs use it to validate S/4HANA business cases, quantify technical debt exposure, and get migration readiness assessments on demand, without waiting weeks for a discovery workshop.

Security and architecture

Is the underlying AI model self-developed or third-party?

C16 uses Anthropic's Claude as the underlying language model. It is not a self-developed model. The differentiation is not the model itself but the SAP-specific tooling, workflows and domain expertise built on top of it: live SAP connectivity, Clean Core classification, code generation and deployment, RAP and CAP builders, process intelligence, and security governance. The model is a component; C16's value is the orchestration layer around it.

Can we choose which LLM to use?

Not at present. The AI engine is standardised on Anthropic Claude across all deployments, so that quality, behaviour and security controls stay consistent for every customer.

Can C16 run entirely within our own environment?

SAP source code and business data queried during a session are processed in memory only and discarded at session end. C16 does not persist your SAP source or business data by default. Connectivity uses a service user that you provision, over a channel you control, either a direct connection to an endpoint you expose or SAP Cloud Connector configured by your team. No inbound access into your network is required, and you keep control of firewall and VPN access. For strict data residency or isolation requirements, C16 can be deployed as a dedicated single-tenant instance inside your own AWS account and region, which confines the platform, the SAP connectivity and all C16-side storage to your AWS boundary. AI inference runs as an external service; the applicable data-processing and cross-border terms are set out in the Data Processing Agreement.

What is the data retention policy for prompts, source code and generated outputs?

ABAP source code, metadata and business data are transient: held in memory only and discarded at session end. Generated documents are retained until you delete them, or 30 days after contract end, whichever comes first. Prompts and conversations are not stored persistently and are held only for session context. No customer prompts, code or data are used to train any AI model, and that is contractually guaranteed. Zero Data Retention can also be enabled with the model provider, so prompt inputs and outputs are not retained after processing.

How are credentials and data secured on your side?

Credentials are encrypted with AES-256-GCM through AWS Secrets Manager, using C16-managed keys with tenant isolation. Audit logs are append-only and tamper-evident, and log redaction automatically strips passwords, API keys and personal data from every entry. Generated documents are retained until deleted or 30 days after contract end. PEOL Technologies, which builds C16, is ISO 27001 certified and SOC 2 Type II compliant, with security controls aligned to ISO 27001 Annex A.8.24.

If generated code causes an issue in testing or production, how is it traced?

Every deployment goes through the standard SAP transport system, so you get a full audit trail inside SAP: transport ID, reviewer, timestamp and change log. C16 keeps its own tamper-evident, append-only audit log of all actions, exportable for compliance. Every workflow includes a consultant review before production deployment, with sign-off recorded. To investigate an issue you can trace back through the SAP transport to a specific change request and cross-reference the C16 audit log for context. Both systems maintain independent trails. Code generated with C16 is supported like any other custom ABAP through your standard change management, and you own the deployed code.

How maintainable is the code long term? Are we locked in?

No lock-in. Versioning is handled by the SAP transport system, which provides native version control, and change management, rollback and transport release all follow your standard SAP practices. Because C16 deploys through SAP's own native mechanisms, a future support team inherits a familiar governance trail: the code is standard ABAP and the deployment path is standard SAP. The C16 audit trail is immutable and exportable, and we run quarterly external penetration testing.