Agentic AI for SAP AMS

Fewer tickets. Faster resolution. A lower cost to serve.

Tessera's resolution agents learn how your client's SAP actually works: their processes, their configuration, their recurring failures. Then they resolve tickets across ServiceNow, SAP, and the business processes behind them, and get sharper with every fix. Built for the providers who run SAP support.

Model the savings See the agents work

Specialized agents · human approval where it matters · every action on the record.

Support queue · live open 312resolved 24h 86awaiting you 7

IDoc 51 · partner profile mismatch

INC0048213 · integration agent · reprocessed BD87

resolved

Batch job failure · variant matched

INC0048990 · basis & jobs agent · rerun SM37

resolved

User not authorized · plant 1200

INC0049002 · access agent · role change drafted

awaiting you

Credit block on sales order

INC0048975 · O2C agent · diagnosing

in progress

"How do I run report S_ALR…?"

INC0049010 · knowledge agent · answered + closed

deflected

every action logged · human approval on higher-risk workmedian MTTR 3h 41m

The economics of SAP support

On a support contract, repetitive tickets are pure cost.

A large share of every SAP support queue is access, integration, batch, and how-to work that has been solved before. On fixed-price contracts that volume eats your margin. On bids it inflates your price. And when an experienced engineer leaves, the cost to serve jumps again.

Swipe to explore →

Margin pressure

Repetition eats the spread

Thousands of near-identical IDoc, access, and batch tickets a year, each handled by a person, on a price that was fixed years ago.

Linear scaling

Growth means headcount

More volume usually means more people. Spikes, ramps, and new clients are expensive and slow to staff.

Fragile knowledge

Expertise walks out

Resolution knowledge is tacit. Attrition and transitions reset it, and reopen and SLA-breach rates climb.

For SAP support providers

A lower cost to serve on every contract you run.

Tessera agents absorb the repetitive, well-understood work across your clients' SAP and ServiceNow estates. You keep the margin on fixed-price contracts, sharpen your bids, and grow without staffing in a straight line.

15 to 25%

fewer tickets

prevented and deflected, year 1

30 to 50%

faster resolution

MTTR on the categories agents handle

12 to 18%

lower cost to serve

less handling effort, fewer FTEs for the same SLA

Illustrative modeled ranges, not guarantees or customer results. A paid assessment replaces them with figures from your own data. Full model below.

Win

Sharper bids

Differentiate proposals with agentic delivery and a quantified takeout model from a real assessment.

Keep

Protect fixed-price margin

Agents handle the repetitive volume, so the contract costs you less to run for the same SLA.

Scale

Deliver more, same team

Absorb growth and volume spikes without staffing in a straight line.

Ramp

Transition faster

Agents plus the knowledge graph shorten knowledge transfer when you take a contract on.

Retain

Resilient to attrition

Every validated fix becomes memory that stays when people leave.

Run

One platform, many clients

Operate across your whole book of business, each client fully isolated.

Brand

White-label

Run it as your own. Your brand, your client relationship, your IP.

Prove

Measured, not promised

Model the savings per contract, prove them in a pilot, report what actually happened.

What it does to a contract

Where the saving comes from.

Most of the takeout is the addressable share of the queue: access, integration, batch, and how-to work. Agents resolve or assist on that subset, and the freed effort becomes margin on a fixed-price contract, capacity you redeploy, or a more competitive bid. The model on the right is illustrative. Your assessment produces the real numbers from your own data.

Scales with your book of business: from a boutique SAP support shop running one contract to a global delivery organization running hundreds.

Illustrative model · one support contractyour data → real numbers

Annual tickets on the contract50,000

Addressable by agents (access, integration, batch, how-to)~40%

Ticket volume removed (prevented + deflected), yr 115–25%

MTTR improvement on handled categories30–50%

Net reduction in handling effort, yr 112–18%

Effort reduction → FTEs to run the same SLAfewer

Where it goesmargin · capacity · bid

Illustrative only, not a guarantee or a customer result. "Ticket volume removed" combines proactive prevention of recurring issues (problem management) with self-service deflection. Effort reduction translates to fewer people needed to hold the same SLA, which on a fixed-price contract is margin. Shares exclude work your platforms already handle and any category below a confidence or safety threshold. A paid assessment replaces every figure with your own.

Run SAP support in-house? The same agents and the same savings model apply to internal SAP application-support teams measured against an SLA, not only to commercial providers.

What the agents resolve

Real work off your queue.

Each outcome is delivered by one or more Tessera agents. We start narrow, on high-volume, lower-risk work that is easy to verify, then expand where the evidence supports it.

Assessment

knowledge agent

Ingest history and return a taxonomy, scored automation candidates, a savings model, and a roadmap. Read-only.

Triage and routing

triage agent

Classify, prioritize, and route to the right team with related-ticket and knowledge suggestions.

Support copilot

diagnosis + knowledge agents

Grounded answers and resolution plans over your tickets, knowledge, and SAP context, with citations.

Integration auto-heal

integration agent

Diagnose and, on approval, reprocess failed IDocs and batch reruns; then validate, document, and close.

Access and authorization

access agent

Compare roles to templates and draft compliant change requests for approval, then apply them through IAM.

Meet the agents that do this →

More than support

The run is where your clients' AI journey starts.

Your clients are under pressure to show an AI agenda. The support estate is the most pragmatic place to start, because it is measurable, it is yours to operate, and it pays for itself. Tessera turns that run into a foundation for the bigger mandate you are already advising them on.

Fund it

Savings pay the way

The cost taken out of run frees the budget your clients need for the rest of their AI program, without a new line item.

Found it

A live process map

Every ticket the agents resolve maps where processes actually break and what they cost. That knowledge graph becomes the starting point for automation beyond support.

Lead it

Expand the mandate

From SAP support into integrations, process automation, and analytics on one governed platform, with you positioned as the partner who delivered the first measurable win.

Tessera is the first step in a broader agentic platform. As your clients' AI journey widens, the same foundation extends into adjacent operations rather than starting over with a new tool.

Why these agents are different

They know your client's SAP, not SAP in general.

Generic IT bots stop at the ticket text. Tessera's agents understand the functional processes underneath: the order-to-cash flow, the configuration, the recurring failure patterns specific to that client. See how it works →

Customer-specific

Your processes, your config

Functional knowledge of one client's SAP, not a generic playbook.

Learned in the assessment

Mapped to a knowledge graph

The assessment builds the graph that links tickets to SAP processes and KPIs.

Compounding

Sharper with every fix

Each validated resolution becomes memory, so the agents keep getting better.

Where the agents run

On your clients' stack, not instead of it.

Their service desk stays their service desk

Agents pick up work from ServiceNow and post proposed resolutions back for approval. No rip and replace; you deploy on top of what the client already runs.

SAP stays the system of record

Agents diagnose and act through supported SAP interfaces, and through sanctioned agent-to-agent paths as they arrive, within the client's authorization model and segregation-of-duties rules.

Start with the data

See what the agents would take off your contracts.

Request an assessment. We scope it with you, ingest historical data under your controls, and return a quantified savings model and a roadmap. No production access.

Model the savings See the agents work

How it works

Learn the client's SAP, then resolve it safely.

It starts with an assessment that maps how that client's SAP actually works into a knowledge graph. The agents reason over that graph, and every action they propose is gated by policy before anything happens. Tessera federates across your systems rather than replacing them.

01 · Integration

Connectors to SAP (S/4, BW/4, CPI and BTP, IDoc, OData, RFC and BAPI), ServiceNow, and identity. Data is queried where it lives; copied only when necessary, within residency controls.

02 · Knowledge

A SAP-AMS knowledge graph and semantic layer, per-client resolution memory, and retrieval indexes that ground every recommendation in your own history.

03 · Agent runtime

Config-driven agents orchestrated through supervisor and graph patterns, with a risk-tiered action model and human-in-the-loop approval queues.

04 · Experience & control

Command center, copilot surfaces, assessment reports, governance and audit dashboards, and cost attribution.

Knowledge graph

A ticket carries its full context.

Incidents, IDocs, orders, configuration, partners, plants, error signatures, and business KPIs live in one semantic model. So when a delivery block arrives, an agent already knows the order, the partner, the interface, the recurring signature, and the KPI it threatens, instead of rediscovering it.

Built during the assessment: 46 entity types and 50+ relationships, plus per-client resolution memory that compounds with every fix.

representative graph fragment · O2C credit-block signature

Neuro-symbolic governance

An agent proposes. Policy decides.

A language model is good at reading a messy ticket and reasoning toward a fix. It is the wrong thing to trust with a production change on its own. So every action an agent proposes passes through a deterministic policy engine before anything happens.

Reason

Agent proposes

The agent reads the ticket and graph context and proposes a specific action with a confidence score and cited evidence.

→

Evaluate

Policy engine

Deterministic rules check it:

action on the allowlist
segregation of duties
environment permitted
confidence above threshold

→

Gate

Decision

Auto-resolve, require human approval, or block and escalate, by risk tier. Every decision is recorded with its inputs and the rules applied.

The policy engine is deterministic and is not overridable by model output or prompt content.

Governance

Safe to put in your client's production.

Action is gated by risk. Higher risk means more oversight, never less. This is what lets a provider put agents near production SAP at all.

Risk	What the agent does	Examples
Low	Auto-resolve within allowlist	Password reset, known IDoc reprocess
Medium	Act and log	Job restart, queue clear
High	Require human approval	Role change, credit release
Critical	Block and escalate	Production config, security grants

Human-in-the-loop approval queues, with shadow mode before any execution
Role- and attribute-based access, with segregation of duties enforced
Source attribution and confidence on every recommendation
Immutable audit of every agent decision and human approval
Per-agent and global rollback and kill switches
Per-client data isolation; one client's knowledge is never shared with another

The runtime

What makes agents safe to deploy.

Swipe to explore →

Connectivity

Reaches your systems

Federated access to SAP (S/4, CPI and BTP, IDoc, OData, RFC, BAPI), ServiceNow, and identity, querying data where it lives.

Memory

Knowledge that stays

Every validated resolution becomes reusable, versioned, and attributable, isolated to each client tenant.

Execution

Acts within limits

Calls SAP via APIs, BAPI, and OData, with RPA only where no interface exists, never beyond the allowlist.

Deployment

Where you need it

SaaS, private cloud, or fully isolated for regulated estates, one control surface everywhere.

Meet the agents → See the product

The agents

A fleet of resolution agents that know your SAP.

Each agent owns a slice of the queue and carries functional knowledge of that client's SAP: the processes, the configuration, the recurring failures. They work across ServiceNow, SAP, and the business processes behind a ticket; act only within the scope you grant; and hand off to a person the moment risk or uncertainty crosses your line.

Swipe to explore →

Triage

read-only

Reads every incoming ticket, classifies and prioritizes it, and routes it; flags duplicates and related tickets.

acts: classify · route · enrich

O2C diagnosis

read-only

Works credit blocks, delivery blocks, pricing and ATP errors. Traces an order to its condition record and credit check.

reads: VA03 · VKM3 view

P2P diagnosis

read-only

Works PO problems, invoice blocks (MIRO), and payment failures, down to the blocking condition.

reads: ME23N · MIR4

Integration auto-heal

approval-gated

Diagnoses failed interfaces and IDocs, proposes a reprocess with corrected data, and executes it once approved.

acts: BD87 · WE20 check

Access & authorization

approval-gated

Reads a user's roles, compares to a template, drafts a compliant change, and applies it through IAM after approval.

acts: SU01 · PFCG draft

Basis & jobs

approval-gated

Handles batch-job failures and queue issues, proposing a rerun with the known variant.

acts: SM37 rerun

Guardrail

policy

Runs every proposal through the policy engine: allowlist, segregation of duties, environment, confidence, risk tier.

acts: approve · gate · block

Verification

read-only

Confirms the fix worked (for example IDoc status 53) or escalates if results look inconsistent.

reads: status · downstream

Knowledge

read-only

Retrieves prior resolutions and articles with citations, and writes validated fixes back to memory on approval.

acts: retrieve · cite

How far they go

The agents earn autonomy. You set the ceiling.

Every agent starts by watching and advising. It moves up only on the categories where it has proven itself, and only as far as you allow. Most estates keep the higher rungs to a short, explicit allowlist.

Swipe to explore →

01

Observe

Learns the queue and builds the picture: taxonomy, clusters, candidates.

02

Recommend

Suggests routing and resolutions grounded in your history, with citations.

03

Assist

Runs read-only diagnostics in SAP and drafts the plan and the ticket response.

04

Execute with approval

Carries out reversible actions only after a person approves, then validates.

05

Resolve within policy

Closes approved low-risk categories end to end, inside the limits you set.

See how it works →

Product

Watch the agents resolve.

Three views: the command center where the fleet runs, a resolution waiting for a human to approve it, and a sample assessment report.

These are Tessera's product interface shown with representative data. They reference SAP transactions (for example BD87, WE20, SU01, SM37) and ServiceNow incident records so they are credible to practitioners. They do not reproduce SAP or ServiceNow interfaces, screens, or logos.

01 · Command Center · the fleet at work

Command Center

client: Northwindenv: PRD-EUwindow: 24h

312

Open incidents

86

Agent-resolved 24h

7

Awaiting approval

3h 41m

Median MTTR

9

SLA at risk

4.2%

Reopen rate

Ticket mix

Access & environment31%

Integration / IDoc / batch19%

Functional O2C / P2P / RTR38%

Technical / performance12%

Agent fleet

Triageonline · 1,204

Integration auto-healonline · 318

Access & authz3 pending

O2C diagnosisonline · 261

Basis & jobsrunning · 44

Activity

09:42INC0048213 · IDoc 51→53 reprocessed (BD87)closed

09:39INC0049002 · role change drafted, sent to approverpending

09:31INC0048990 · batch rerun SM37, variant matchedrunning

09:27INC0048975 · credit block, routed to SAP O2C L2routed

Resolution funnel · 24h

418

Detected

402

Diagnosed

131

Proposed

124

Approved

121

Validated

118

Closed

02 · An agent's proposal, waiting for you

Proposed resolution · access & authorization agent

INC0049002SAP Security L2P3

Incident

short descUser cannot create sales orders for plant 1200, "not authorized"

reported bym.chen (Order Mgmt)

sourceServiceNow incident

What the agent found

root causeMissing authorization object on role; SU53 shows failed check for plant 1200

matched signatureAUTH-O2C-PLANT (412 prior tickets)

confidence0.88

What the agent will do · reversible

Compare m.chen's roles to the "Order Management EU" template (PFCG)
Draft role change: add the missing plant-1200 authorization
On approval, submit change request to IAM, no direct production edit
Verify the failed check now passes; update and close the incident

Risk & controls

HIGH · requires your approval

Preconditions the guardrail agent checked

✓ Segregation of duties: no conflict introduced
✓ Action on allowlist for this tenant
✓ Change routed through IAM, not direct edit
✓ Rollback: revert role assignment

Your decision

assignedJ. Okafor (Security lead)

SLAapprove within 4h

Approved by J. Okafor · agent submitted change request to IAM · verifying… ✓ failed check now passes · incident closed.

audit: proposal, approval, execution, verification, closure recorded · rollback available

03 · Assessment report

AMS Assessment · Summary

period: trailing 12 months · source: your service desk + SAP support data · status: draft

read-only · no production access

48,910

tickets analyzed

41%

recurring-pattern share

12–18%

modeled effort reduction yr 1

3.9h

median MTTR (current)

Top repetitive issues

Error signature	Volume / yr	Median MTTR	Modules	Candidate agent / playbook	Conf.
IDoc 51 · partner mismatch	3,410	2.1h	O2C · EDI	Integration agent · BD87 + WE20	0.93
AUTH · plant not authorized	2,980	5.4h	Security	Access agent · template diff + IAM	0.88
Credit block (VKM)	2,240	3.0h	O2C · FIN	O2C diagnosis · diagnose + route	0.74
Batch job failure · variant	1,760	1.6h	Basis	Basis & jobs · rerun SM37	0.81
Invoice block (MIRO)	1,510	4.2h	P2P · AP	P2P diagnosis · diagnose + draft	0.69

Recommended roadmap

Phase 0

Assess

This report. Validate candidates and savings against your data.

Phase 1

Copilot

Triage and grounded resolution suggestions in shadow mode.

Phase 2

Approved execution

IDoc reprocess, batch rerun, access assist, with approval.

Phase 3

Closed-loop

End-to-end resolution for approved low-risk categories.

All figures are modeled from your own data and are targets, not guarantees. The effort-reduction range reflects the share that is both high-volume and low-risk after excluding work your platforms already handle and any category that fails a confidence or safety threshold. Data-quality findings (for example 6% of tickets with no resolution notes) are listed in the full report.

See it on your data

The mockups are generic. Your assessment would not be.

Request an assessment and we will run this against your own history.

Model the savings How it works

Legal

Privacy Policy

Last updated: June 2026

This Privacy Policy explains how NuScale Technologies ("we", "us") handles information in connection with the TesseraOps website and product enquiries. We keep data collection to the minimum needed to respond to you and operate the site.

Information we collect

Information you provide. When you email us (for example, to request a savings model), we receive your name, email address, and anything you choose to include.
Usage data. Our hosting provider may log standard technical data such as IP address, browser type, and pages visited, for security and aggregate analytics.

How we use it

To respond to your enquiry and provide the information or assessment you requested.
To operate, secure, and improve the website.
To comply with legal obligations.

We do not sell your personal information.

Sharing

We share information only with service providers who help us run the site and communications (for example, hosting and email), and where required by law. These providers are bound to use the information only to provide their services to us.

Data retention

We retain enquiry correspondence for as long as needed to address your request and for reasonable business and legal record-keeping, then delete or anonymise it.

Your choices

You may request access to, correction of, or deletion of the personal information you have shared with us by emailing info@nustudio.ai.

Contact

Questions about this policy? Email info@nustudio.ai.

Legal

Terms of Service

Last updated: June 2026

These Terms govern your use of the TesseraOps website operated by NuScale Technologies. By using this site you agree to these Terms.

Use of the site

You may use this site for lawful, informational purposes. You agree not to misuse the site, attempt to disrupt it, or access it in a way that violates applicable law.

No warranties

The site and its content are provided "as is" without warranties of any kind, express or implied. We do not warrant that the site will be uninterrupted, error-free, or secure.

Illustrative figures

Any savings, outcome, or performance figures shown on this site are illustrative models for discussion only. They are validated per contract and are not guarantees of results or representations of any specific customer outcome.

Intellectual property

The TesseraOps name, logo, content, and product screens are the property of NuScale Technologies. Product screens are our own interface with representative data and do not reproduce any third-party vendor's interface or logos. References to SAP, ServiceNow, and other marks are for identification only and remain the property of their respective owners.

Limitation of liability

To the maximum extent permitted by law, NuScale Technologies will not be liable for any indirect, incidental, or consequential damages arising from your use of the site.

Changes

We may update these Terms from time to time. Continued use of the site after changes constitutes acceptance of the updated Terms.

Contact

Questions about these Terms? Email info@nustudio.ai.