SELECCIÓN AAS is our AI product that runs SAP support autonomously — AI agents resolve routine tickets in minutes, with certified consultants owning escalations and architecture. 70% of tickets resolve with no human in the loop — an outcome a human AMS team structurally cannot deliver, no matter how many people you add or which vendor runs it. It runs on SAP ECC and S/4HANA alike.
The Hard Truth
Hiring more people or switching AMS vendors does not fix a model that is structurally limited by human capacity.
Whether you outsource or run it in-house — the model most companies use today has four built-in problems.
!
2–3 Days
For a routine SAP ticket — an access request, a report error, a config tweak. That wait is a structural feature of human AMS, not a bug.
!
Zero
Resolution at 2am on a Sunday. Human teams have shifts and timezones. AI agents don't have office hours.
!
Lost
Every time a key resource leaves. AMS knowledge lives in people's heads — every departure is a service risk.
!
High Cost
For largely repetitive work. Most SAP AMS tickets follow documented patterns, priced at senior resource rates.
Not Improvements — Outcomes Out of Reach for Any Human AMS Team
50 Tickets, Simultaneously
A human team queues — tickets stack up at month-end close. An AI agent handles all 50 at once, no matter the volume.
Same Response, 2am or Noon
No shifts, no timezones, no holidays. A ticket raised at 2am resolves in the same minutes as one raised at noon.
Full Knowledge From Day One
A new hire takes 3-6 months to learn your landscape. SELECCIÓN AAS ingests 12-24 months of ticket history before ticket one.
Knowledge That Compounds
Human teams forget, and knowledge leaves when people do. AAS's knowledge base only grows — turnover changes nothing.
Minutes, not the standard 2-3 day SLABest SAP people back on projects, not L1 ticketsSLA compliance guaranteed, full audit trail every time
What Makes SELECCIÓN AAS Different
Every vendor now claims "AI resolves tickets." Here's what makes ours true.
Built Agentic From Day One
Not a chatbot bolted onto a ticketing tool, and not RPA scripts wrapped in a UI. Each agent is architected ground-up with its own identity, scope, and audit trail.
The Only Native Execution Engine in the World
AAS embeds Enterprise MCP's execution engine directly — a SAP-certified add-on running natively inside your ABAP stack, reaching BAPIs, transactions, and tables the way your own developers do.
No API, Version, or Service Pack Dependency
Works natively across SAP ECC and S/4HANA, at any enhancement or service pack level. Nothing to re-integrate every time you patch or upgrade.
Scales Across Every Customization — Zero New Code
Handles your custom Z-transactions, custom BAPIs, and configuration exactly as they exist today.
Dual-Mode LLM — Public to Learn, Private to Run
Learning mode can use a large public LLM (Claude, Gemini) — Enterprise MCP shares only ticket metadata, never business data. Once live, it switches to a small, self-contained model hosted inside your environment. Nothing leaves your infrastructure.
Every action executes within the acting user's existing SAP authorization profile — the agent can never exceed what that role permits. Role-based access control stays the single enforcement point, with every action logged and attributable, end to end.
From Ticket-History Ingestion to Full Autonomy
Public LLM · Metadata Only
1. Learn
SELECCIÓN AAS ingests 12-24 months of ticket history
Environment, landscape & integration discovery
Classification models trained before ticket one
Baseline SLAs and escalation paths agreed with your team
Private LLM · Self-Hosted
2. Human-in-the-Loop
AI proposes resolutions; your team reviews & approves
Confidence scores rise with every validated match
Autonomous coverage expands ticket type by ticket type
Full transparency — nothing goes live unsupervised
Private LLM · Self-Hosted
3. Autonomous
AI resolves L1/L2 tickets independently, in minutes
Your team owns escalations, architecture & P1 root cause
Knowledge compounds — every resolution improves the next
100% audit trail, continuous SLA compliance
Running Support During Your S/4HANA Migration
The questions we hear most from teams migrating off ECC.
Most AMS vendors now claim "AI-powered" support. Few can show you what percentage of tickets actually resolve without a human, or give you a real audit trail proving it. The question isn't whether your current vendor uses AI somewhere — it's whether they can show you a number, and whether that AI executes natively inside SAP or sits behind an API layer that adds latency and risk. We can show you both. If your current vendor can't, that's the gap worth testing.
SELECCIÓN AAS runs support independently of your migration team, so your best SAP people stay on the migration instead of getting pulled into L1 tickets. It's built for exactly this — freeing internal capacity during the highest-demand phase of a program.
Yes. AAS can take on full ECC support ownership, including tickets your internal team would otherwise have to context-switch into, while your migration team stays focused on the S/4HANA build.
Both, without switching models. AAS runs on ECC and S/4HANA the same way, so you're not standing up a new support approach partway through your migration.
AAS continues running during cutover and hypercare, with human escalation paths in place for the higher-stakes issues that come up during go-live specifically. This is a phase where human oversight matters most, and our model keeps humans at that checkpoint.
Yes. If your migration is phased, AAS can support both environments in parallel, since it isn't tied to a single system's architecture.
AAS ingests 12-24 months of your ticket history during onboarding before handling a single live ticket, so it starts with landscape knowledge already in place rather than a slow ramp-up. Exact timelines depend on your ticket volume and landscape complexity — this is something we size specifically for your environment.
It's a fair instinct to assume adding AI during a migration adds risk. In practice, it's the opposite: AAS starts in a human-in-the-loop mode, where AI proposes resolutions and your team reviews and approves every one, with autonomy expanding only as confidence rises ticket type by ticket type. Nothing goes live unsupervised, and the full audit trail gives you more visibility into support activity during the migration, not less.
AAS ingests 12-24 months of your ticket history, along with your environment, landscape, and integration details, before classification models are trained and any live ticket is handled.
Your team owns escalations, architecture decisions, and P1 root cause investigation — the judgment-heavy work stays with certified consultants. AAS handles the documented, repeatable patterns that make up the majority of ticket volume, not the exceptions.
The intent is redeployment, not reduction. Freeing your best SAP people from L1 ticket volume means they're available for the migration, for optimization work, or for the next project — not idle.
Traditional AMS pricing is largely driven by headcount and seniority mix. AAS pricing reflects AI agents handling the repetitive volume, with human oversight priced for the judgment-heavy work — which is what drives the cost reduction relative to a fully human-staffed model.
Why Run Your SAP AMS With SELECCIÓN Consulting, Now?