AI on your SAP data, without the CPI message bill.
QSAP connects the AI your teams already use to SAP S/4HANA Cloud through SAP's standard released APIs. AI questions add no Integration Suite messages, and QSAP itself is a flat fee: you bring your own AI, and we never resell or mark up model usage. Every person signs in with SSO, sees only what their SAP authorisations allow, and leaves a tamper-evident audit record.
- Lower costNo per-message metering in the AI path, and leaner payloads for the model
- Bring your own AIClaude, ChatGPT, Gemini and other assistants, for one flat fee
- SecureSSO and per-user SAP authorisation on every call
- AuditableA tamper-evident record of who asked what
Built by Quantana for manufacturers on S/4HANA Cloud and SAP BTP. Illustration uses demo data.
Why AI makes CPI expensive
SAP Integration Suite is priced per message. That works for scheduled interfaces with a predictable volume. AI traffic is the opposite: many small calls, and more of them every time another team starts asking questions.
Every message is metered
Integration Suite bills in blocks of messages. A request and its reply count as one message up to 250 KB combined, but large payloads, retries and multi-step flows (API Management in front of an iFlow, splitters) multiply the count. Every environment you run, from development and quality to production, adds its own volume.
One question is several calls
An AI agent rarely answers in one lookup. It lists a summary, fetches the detail, then checks a related record, typically 4 to 8 tool calls per question. So message volume multiplies, and the bill grows with adoption.
QSAP takes the meter out of the AI path
QSAP calls SAP's released APIs directly, so AI questions add no Integration Suite messages. You pay a flat monthly fee, and it runs in your own SAP BTP account or fully managed by us. It also sends the AI only the data it needs, which cuts AI token spend.
| Aspect | AI routed through CPI | AI through QSAP |
|---|---|---|
| Pricing model | Metered per message (or drawn from a capacity allowance) | A flat monthly fee, with no per-question, per-token or per-message charges |
| Cost driver | Number of messages, multiplied by large payloads, retries and multi-step iFlows | Your package and user count, not question volume |
| As usage grows | Rises with every user, question and tool call | Stays flat within your package |
| Payload sent to the LLM | Whatever the integration passes through, often the full SAP record | Only the fields the question needs, cleaned and cached |
| Environments | Development, quality and production each add messages | No message charges in any environment |
| Governance | Built per iFlow; per-user identity often lost behind a technical user | SSO, per-user SAP authorisation, role policy and a tamper-evident audit trail on every call |
Estimate your CPI cost for AI traffic
All amounts in EUR. Start from list price, then use your own numbers. Nothing is sent anywhere; the sums run in your browser.
- Tool calls per month
- 1,320,000
- CPI messages per month
- 1,320,000
- Estimated CPI cost per month
- €802.56
- QSAP hosting per month
- €9.00
- Message charges avoided, net of hosting, per month
- €793.56
- Per year
- €9,522.72
100 users × 100 questions × 6 tool calls × 22 days = 1,320,000 tool calls. × 1 messages = 1,320,000 messages. 1,320,000 ÷ 10,000 × €6.08 = €802.56 per month. Difference = CPI cost − €9.00 hosting = €793.56 per month, × 12 = €9,522.72 per year.
Estimates only, not your actual cost. The default uses a single list-price tier with no volume discount or included messages. SAP pricing varies by contract, edition (message-based or capacity-based) and region. Validate with your SAP account team. Excludes SAP user licensing, which is unchanged, and the QSAP subscription (see pricing).
Bring your own AI. Pay us a flat fee.
Your AI subscription, your model choice, your data policy. QSAP charges a flat fee, not per question, per token or per message, so costs stay predictable as adoption grows. That is the opposite of a per-message integration meter, where every new user adds to the bill.
Works with the assistants your teams already use
- ClaudeDesktop, claude.ai, Claude Code
- ChatGPTthrough connectors
- GeminiGemini CLI, Gemini Enterprise
- Microsoft Copilot Studiothrough connectors
- Other assistantsthat support standard AI connectors
- QSAP Portalyour own OpenRouter key
Flat monthly fee, from ₹50,000 a month. We don't resell or mark up AI usage. See pricing.
Product names are trademarks of their respective owners. QSAP is not affiliated with or endorsed by Anthropic, OpenAI, Google or Microsoft.
Cost is the first blocker. It isn't the only one.
Governance has no answer for "who asked?"
A shared technical user lets a model read everything that user can read. Auditors want to know which person saw which data, under which authorisation, and whether the log could have been edited.
Raw SAP data is heavy for AI
A typical SAP response carries dozens of fields and technical metadata for every record. The AI pays to read all of it and answers no better for it.
How it works
Your AI assistant connects to QSAP like any other tool. QSAP holds the SAP connection, checks every request, and calls SAP's released APIs as the signed-in person.
Sign in once
Users authenticate through your identity provider when they first connect their assistant. QSAP never sees a SAP password.
Policy before SAP
Role groups decide which tools and plants each person can use. Denied requests stop here and are still recorded.
Call SAP as the person
SAP sees the real person and applies their own authorisations, so the AI never sees more than the person asking.
Answer compactly, log permanently
Responses are trimmed to the fields that matter, and the call is recorded in a tamper-evident audit trail.
Up and running in days, not months
Most of the work is ours. Here is what each side typically does, from first access to first answers.
Your SAP admin, typically about 1 hour
Grant access
Create one communication arrangement in S/4HANA for the released APIs QSAP uses. It is standard SAP configuration: no custom code, no transports, no changes to SAP.
Your IT team, typically about half a day
Connect sign-in
Point QSAP at your existing identity provider, such as SAP Cloud Identity Services or Microsoft Entra ID, and map your teams to roles: who can see which plants and data.
Quantana, typically about 1 day
Deploy
We deploy QSAP into your own SAP BTP account, or run it for you as a managed service. Nothing to install on users' machines.
Each user, typically about 2 minutes
Connect your AI
Add QSAP as a connector in Claude, ChatGPT, Gemini or another assistant your teams already use, then sign in with a work account. See how for each assistant.
Week 1
First answers
The pilot starts with your top reports. QSAP records every request from day one, so usage is visible straight away.
What you don't need
- No CPI iFlows to build or maintain
- No SAP code changes
- No new AI licences: bring your own
- No data copied out of SAP into another warehouse
Security and governance your auditors can check
Designed to SAP's reference architecture for third-party MCP servers: released APIs only, every call authenticated, authorized and audited. Confirm fit with your SAP account team.
- Single sign-on
- Users sign in through your identity provider, such as SAP Cloud Identity Services. Sessions are short-lived and tied to that identity.
- Per-user SAP authorisation
- Either SAP checks each person's own authorisations on every call, or your administrator chooses one technical user and QSAP enforces each person's access.
- Role and plant policy
- Map identity groups to the tools and plants they may use. A plant manager sees plant 1710; a viewer can check stock and nothing else.
- Tamper-evident audit trail
- Every call is stored with the person, tool, SAP request, status and timing. Records are sealed in sequence, so any edit or deletion shows up when the trail is verified.
- Released APIs only
- QSAP uses a fixed set of SAP's released APIs and asks only for the fields each question needs. There is no open-ended query access for an AI to misuse.
- Guarded inputs and rate limits
- Every request is checked before it reaches SAP, and each person has a fair-use limit per minute.
Reads are instant. Changes stay in human hands.
QSAP reads SAP data directly. For supported changes, such as raising a purchase requisition or updating a delivery date, the AI prepares the change and the user reviews and posts it in the standard SAP screen, with the fields already filled in.
- SAP's own validations, authorisations and approval workflows check every change.
- A named person approves each one, and it is recorded in the audit trail.
- No extra per-document SAP charges for system-created documents, as changes are posted by your users under their existing SAP user licences.
Every team gets the answers they need, and only the data they're allowed to see
People ask in plain English, in the assistant they already use. What each person can see is set by your rules, not by the AI.
-
Finance
- “Which customers are over 90 days overdue in the South zone?”
- “Margin by product family this month vs last”
The daily receivables flash and month-end tracking, without Excel dumps.
-
Sales and commercial
- “What's pending dispatch for my top 20 customers?”
- “Billing vs target for my region today”
Live order status in the field, limited to their own region.
-
Production and planning
- “Which orders this week are short of components?”
- “What goes into this product?”
Catch shortages before they stop the line.
-
Procurement
- “Which purchase orders are overdue from suppliers?”
- “Coverage days for our key raw materials”
Chase the right suppliers first. Procurement can have the AI draft a purchase requisition; they review and post it in SAP.
-
Stores and dispatch
- “What can we ship today with stock on hand?”
Plan the day's dispatch from live stock, plant by plant.
-
Marketing and product
- “Which product families grew fastest this quarter?”
- “Product catalogue details for the new range”
Read-only product and sales-volume insight, with no prices, margins or customer balances.
-
Leadership
- “How are we tracking against this month's target?”
One question instead of a stack of reports, plus a savings dashboard.
-
IT and audit
- “Who asked what, when, and with which data?”
A tamper-evident trail of every request, and access managed in one place.
Example questions for a typical manufacturer. Which questions QSAP can answer depends on the business tools set up for your pilot.
Who sees what: an example
- Full access
- Own region or plant only
- No access
| Team | Products and catalogue | Stock | Production | Purchasing | Sales orders | Prices and margins | Customer balances | Vendor payments |
|---|---|---|---|---|---|---|---|---|
| Finance | Full access | Full access | Full access | Full access | Full access | Full access | Full access | Full access |
| Sales | Full access | Full access | No access | No access | Own region or plant only | No access | Own region or plant only | No access |
| Production and planning | Full access | Full access | Full access | Own region or plant only | No access | No access | No access | No access |
| Procurement | Full access | Full access | No access | Full access: Including vendor pricesIncluding vendor prices | No access | No access | No access | No access |
| Stores and dispatch | Full access | Own region or plant only | No access | Own region or plant only | Own region or plant only | No access | No access | No access |
| Marketing | Full access | No access | No access | No access | Own region or plant only: Volumes, no valuesVolumes, no values | No access | No access | No access |
| Leadership | Full access | Full access | Full access | Full access | Full access | Full access | Full access | Full access |
The model reads what it needs and nothing more
Less data per answer means lower AI cost and more reliable answers. QSAP gets there four ways.
- Only the fields that matter. Every SAP request asks for just the data the question needs.
- Clean answers. SAP's technical wrapping is stripped before the AI sees anything.
- Smart caching. Repeat questions are answered from a short-lived cache that respects each person's access, so one person's results never reach another.
- Summary, then detail. The AI lists the open orders first and fetches full records only for the ones that matter.
{"d":{"results":[{"__metadata":{"id":"…/A_ProductionOrder_2('1000231')",
"uri":"…","type":"API_PRODUCTION_ORDER_2_SRV.A_ProductionOrder_2Type"},
"ManufacturingOrder":"1000231","ManufacturingOrderCategory":"10",
"ManufacturingOrderType":"YBM1","OrderIsCreated":"","OrderIsReleased":"X",
"OrderIsPrinted":"","OrderIsConfirmed":"","OrderIsPartiallyConfirmed":"",
"OrderIsDelivered":"","OrderIsDeleted":"","OrderIsPreCosted":"X",
"MfgOrderCreationDate":"\/Date(1790380800000)\/", … 80 more fields,
"to_ProductionOrderComponent":{"__deferred":{"uri":"…"}}, …{"id":"1000231","material":"FG-DIV-200","status":"released",
"short":[{"component":"CART-35","need":40,"available":12}]}CPI vs QSAP: the benchmark
We ran the same 11 everyday questions three ways: through a CPI-style pass-through, through CPI with its requests trimmed to the needed fields, and through QSAP.
- 0CPI messages. QSAP connects to SAP directly, so nothing is metered through Integration Suite.
- 94%fewer AI tokens than a pass-through that sends full SAP records. Even against a well-tuned CPI flow, it's 83% fewer.
- 69%less data pulled from SAP, because QSAP asks only for the fields the answer needs.
- 2 in 11answers came straight from cache, with no SAP call. Every answer shows when its data was fetched.
Simulation on S/4HANA-compatible demo data. Results vary with your questions and data. We'll run it on your own system during the pilot.
A planner's question, answered in one turn
The assistant picks the tools, QSAP checks the planner's role and runs the calls, and the answer arrives with the order numbers a planner can act on. Illustrative example on demo data for a bath-fittings manufacturer.
Production planner
Which production orders in plant 1710 are short of components?
Open orders found 6 released orders in plant 1710
Shortage check run for each of the 6 orders
Assistant
Two of the six released orders in plant 1710 are short:
| Order | Product | Short component | Need | Available |
|---|---|---|---|---|
| 1000231 | Diverter, 3-way | Ceramic cartridge 35 mm | 40 | 12 |
| 1000237 | Basin mixer | Aerator M24 | 150 | 90 |
The other four orders have full component cover. Want me to check the cartridge stock in plant 1720?
Runs where your SAP landscape already lives
QSAP is a single service with a small footprint. Your data travels between SAP, QSAP and the model provider you choose, and the audit trail stays in your environment.
| Option | Good fit when | Notes |
|---|---|---|
| SAP BTP, Cloud Foundry | You already run extensions on BTP | Deployed into your subaccount next to your other BTP apps |
| SAP BTP, Kyma runtime | Your team standardises on containers within BTP | Runs as a standard container |
| Your own container platform | You prefer your own cloud or data centre | A standard container for your cloud or data centre |
| Managed by Quantana | You'd rather not run it yourselves | We host and operate it for you, from ₹10,000 a month on top of your package |
What's next
The same tools, policy and audit trail, opened up to more people and more ways of working.
QSAP Portal
A web chat for people without an AI assistant licence. Bring your own model key through OpenRouter and use any supported model, with the same SSO, policy and audit.
Dashboards
Save the answers you ask for every week as dashboards that refresh from SAP.
Report downloads
Export any answer as CSV, Excel or PDF, with the audit reference attached.
Questions IT leaders ask
How does QSAP affect our SAP licensing?
QSAP works within your existing SAP user licensing: each request runs as a named SAP user or as a technical user your administrator chooses. Commercial terms differ between contracts, so validate how AI-driven access is treated with your SAP account team.
Where does our data go?
QSAP runs in your BTP subaccount or your own infrastructure, so you choose the region. SAP data passes from SAP to QSAP and then to the model provider your users connect with. The audit trail is stored where QSAP is deployed. Pick a model provider and region that meet your data-residency rules.
Which SAP versions does it support?
QSAP is built for SAP S/4HANA Cloud and its released APIs, covering stock, production orders, bills of material and sales orders. If you run S/4HANA Cloud Private Edition or on-premise with the same released APIs exposed, talk to us about your landscape.
Do we need SAP Integration Suite?
Not for QSAP. It calls released APIs directly, so AI traffic does not consume Integration Suite messages. Your existing integrations can keep running on Integration Suite as they do today.
Can QSAP change data in SAP?
QSAP reads SAP data directly. For supported changes, such as a purchase requisition or a new delivery date, the AI prepares the change and the user reviews and posts it in the standard SAP screen with the fields pre-filled. SAP's own validations, authorisations and approval workflows apply, a named person approves every change, and each one is recorded in the audit trail. Because your users post the documents, there are no extra per-document SAP charges for system-created documents. Which changes are supported is agreed in your pilot; validate commercial terms with your SAP account team.
Which AI assistants work with it?
Claude, ChatGPT, Gemini, Microsoft Copilot Studio and other assistants that support the Model Context Protocol (MCP), the open standard for connecting AI assistants to business tools. You keep your own AI subscription; QSAP is a flat fee. The upcoming portal will let teams use other models with their own OpenRouter key.