Home · Enterprise
Enterprise · On-premise · Campus compute

Deploy WinGrants AI where procurement needs it to run.

For institutions that need more than managed software. Choose Enterprise Lite for on-premise application and data with approved inference, or Enterprise Campus Compute for fully local model inference on your own infrastructure.

Book a demo → See what's included
From €30k / year + €5k setup. Annual licence · uncapped use across every EU program.
✓ISO/IEC 27001:2022 certified ✓On-premise available ✓Air-gap capable ✓Single sign-on ✓Audit logs
Choose your path

Two Enterprise deployment paths.

Enterprise is not one deployment. It is a choice between Enterprise Lite for on-premise data with approved inference, and Campus Compute for fully local model inference.

Enterprise Lite

On-premise application + approved inference

Run the WinGrants AI application, database, files, identity layer, and audit logs in your own infrastructure or approved European cloud. Model inference routes only to approved endpoints defined in the agreement. This is the faster path when your institution needs local data control but does not have local graphics processing units.

  • Application and data in your environment
  • Approved model endpoint
  • No local graphics processing units required
  • Institutional login available
  • Audit logs and role-based access controls
  • Typical rollout: 3–5 weeks
Enterprise Campus Compute

Fully local model inference

Run the application and model inference inside your institutional perimeter. Model calls route through local vLLM servers using open-weight models on your graphics processing units or high-performance computing environment. This is the high-control path when procurement requires zero runtime third-party model calls.

  • Application and data in your environment
  • Local open-weight models
  • No runtime third-party model calls
  • Version-pinned model stack
  • Air-gap capable
  • Typical rollout: 4–8+ weeks, depending on compute and network approvals
Who Enterprise is for
UniversitiesCentralised grant office · multi-faculty
RTOs & institutesResearch orgs with consortium IP duty
National agenciesSovereign deploy · classified data fit
Industrial AssociationsFederation-wide rollouts · member confidentiality
Why this conversation exists

Some proposal data cannot go through standard managed AI.

If you're reading this page, you've already tried to procure something. 3 things keep blocking the sign-off.

01
Procurement

Your institution needs a deployment it can inspect and approve.

EU research institutions and national agencies have a duty of care over consortium IP. For sensitive consortium work, procurement may require local data storage, approved subprocessors, institutional identity, audit logs, or fully local inference — before the conversation even reaches the grants office.

The blockRisk register · data residency · sub-processor lists
02
Model policy

Model providers need to match your risk posture.

Some institutions approve commercial models for quality. Others require open-weight models, European providers, or local inference. Enterprise lets the model path be scoped instead of assumed.

The blockProvider control · Model risk posture · Approval needed
03
Multi-tenant risk

Your consortium intellectual property may require isolation.

A €4M EIC proposal is multiple partners' work. Putting it through a multi-tenant database that other organisations also use is a category of risk most consortium leads can't get signed off — rightly so.

The blockPartner agreements · DPA chains · audit trails
The shape of the decision

Managed AI vs Enterprise self-hosted.

Standard managed AI

Someone else's cloud, model and price.

WinGrants AI Enterprise — your deployment, your controls, your model policy

Your deployment, your controls, your model policy.
Residency
"EU region" promiseSub-processors elsewhere
Application and data hosted in your environment or approved cloudEnterprise Lite or Campus Compute
Models
Closed-sourceVendor decides when they change
Approved endpoint or local open-weight modelsVersion-pinned where required.
Keys
Vendor holds the keysEncryption + API keys
Customer-approved key and routing policyBYOK supported
Pricing
Per-seat or per-tokenClimbs over time
Fixed annual licenceNo per-seat or per-token climb
Exit
Data export onlyNo model, no workflows
Database export, prompts, scorers, workflow configuration, and deployment artifacts under contractYours under the agreed export package
Same engine, three relationships

Is Enterprise actually what you need?

Most teams don't need on-prem. Here's the honest read on where each plan fits — so you don't over-buy.

If you're a solo PI

Self-service

One researcher, one proposal at a time. EU-hosted SaaS, no procurement gate, pay only when you commit to submitting.

€1,000/ proposal
European Union-hosted application, model choice toggle, no subscription No subscription · no annual minimum Credit Replacement Guarantee on your first credit
See Self-service →
If you're a research team

AI Enablement

3–30 researchers in a grants office, RTO, or consultancy. Fixed annual licence; single-tenant workspace, private workflow customization, configurable model routing.

From €15k/ year · +€5k setup
Single-tenant workspace, European Union-hosted 4-week founder-led setup · configurable model routing + isolated database Uncapped use across every EU programme
See AI Enablement →
If procurement says SaaS is out

Enterprise · you are here

Institutions that need a deployment they can inspect and approve. On-premise or approved cloud for Enterprise Lite; fully local open-weight models for Campus Compute.

From €30k/ year · +€5k setup · scoped by deployment
Enterprise Lite — on-premise application and data with approved inference Campus Compute — fully local open-weight models on your infrastructure Founder-led deployment; founders in the Slack channel
Book a demo →
Not sure? The fastest way to find out is a 30-minute discovery call. If Enterprise isn't the right fit, we'll route you to AI Enablement — same engine, simpler deployment.
Your AI, your servers

Approved model routing or fully local open-weight models.

Model policy matched to procurement risk.

Enterprise supports two model paths. Enterprise Lite routes inference to approved model endpoints defined in the contract. Enterprise Campus Compute runs open-weight models locally through vLLM on your own infrastructure. For fully local deployments, models are version-pinned and upgrades require approval.

  • Enterprise Lite: approved sovereign or customer-approved inference endpoint.
  • Campus Compute: local vLLM inference with open-weight models.
  • Version-pinned models where required. Predictable behaviour over multi-year cycles.
  • No silent provider changes. Routing policy is documented before production use.
  • No customer proposal content used to train WinGrants AI.
wingrants.internal.yourorg.eu
CAMPUS COMPUTE MODE
Local inference — no runtime third-party model calls
Reasoning modellocal
Drafting modellocal
Multilingual modellocal
Long-context modellocal
Fallback modellocal
NO RUNTIME MODEL EGRESSGDPRISO 27001DPA
Deployment & security posture

Your perimeter. Your keys. Your audit trail.

Built for institutions where data sovereignty isn't optional.

2 Enterprise deployment topologies — Enterprise Lite and Enterprise Campus Compute. Pick the one that matches your IT department's risk posture. Identity, encryption, and audit logging integrate with the stack you already run. Managed single-tenant cloud belongs under AI Enablement, not Enterprise.

  • 2 Enterprise deployment topologies. Enterprise Lite (on-premise application and data with approved inference) or Campus Compute (fully local inference, air-gap capable).
  • ISO/IEC 27001:2022 certified, GDPR & DPA aligned. Independently audited, continuously monitored.
  • AES-256 at rest, TLS 1.3 in transit. Private keys never leave your environment; BYOK supported on Enterprise.
  • SSO & SCIM with Azure AD, Okta, eduGAIN, Shibboleth, Keycloak — your institutional IdP, no custom integration.
  • Full audit trail, RBAC, per-project isolation. Every action signed and append-only.
On-premise · your infrastructure wingrants.yourorg.local:8443
ISO 27001Independently audited, continuously monitored.
GDPR & DPAEU-hosted, right to erasure enforced.
AES-256 + TLS 1.3End-to-end encryption, zero trust.
SSO / SAML + RBACAzure AD, Okta, SCIM provisioning.
Private cloudOn-premiseAudit logsZero trust
2
Deployment-specific hosting — Enterprise Lite and Campus Compute are scoped to your approved hosting and model-routing requirements
0
Runtime third-party model calls in Campus Compute
4h
Incident notification SLA
2–4wk
Ramp-up to production
What you get

More than software. Deployed to match your procurement path.

Enterprise isn't a bigger licence — it's a different relationship. Your IT team gets a deployment they can audit, on infrastructure they control. Your grants office gets the same hands-on rollout as AI Enablement. And you get the founders behind both.

01

Self-hosted platform

The full WinGrants engine deployed in your approved environment. Enterprise Lite keeps application data in your environment and routes inference to approved endpoints. Campus Compute keeps inference local as well. Unlimited seats across every PI, grant officer and project manager.

Air-gap capable · unlimited seats
02

Onboarding & team training

2–4 week founder-led ramp-up with your IT and grants office in the room. We co-deploy with your engineers, configure security policies live, and train your researchers on real cases before we leave.

IT + grants office · real cases
03

Model setup

Enterprise Campus Compute runs open-weight models locally through vLLM, pinned to approved versions. Enterprise Lite uses approved model endpoints defined in the agreement. No provider changes are made without approval.

Pinned models · approval before changes
04

Direct line to the founders

Shared Slack channel with the founding team for both your IT and your grants office. Async, no tickets, no support tiers — including roadmap input. Quarterly tuning sessions and ongoing EU funding advisory included year-1.

Year 1 included · roadmap access
Prefer to meet in person? For RTOs, universities and EU agencies, we travel to your campus — security review, kickoff, or a working session with your grants office. No travel charge for EU institutions.
Arrange an on-site visit →
Ali Haider, Co-Founder of WinGrants AI
Why we built Enterprise
Universities and RTOs kept telling us — we love the platform, but our procurement team needs a deployment path that matches our data and model-risk policy. Enterprise is the answer to that conversation.
Ali Haider
Co-Founder & CTO at WinGrants AI
ROI calculator

What would Enterprise unlock?

Personalise the inputs to match your institution. €30k/year licence anchor + €5k setup (scoped on the call) — uncapped use under the selected Enterprise deployment.

Assumptions

  • Annual licence anchored at €30k/year + €5k one-time setup for the calc — actual price scoped to org size, integration scope, SSO/audit complexity
  • Self-hosted on your infrastructure — writing time per proposal drops by up to ~70%
  • Uncapped use on your servers, no per-proposal fee
  • External consultancy spend drops to zero
  • Success rate is held flat at your historic rate (conservative)
Faster
Annual time freed iApples-to-apples: hours saved if you ran the SAME number of proposals at Enterprise's pace. Formula: p × (h − hnew).4,620h
Hours per proposal220h → 66h
Better
Proposals submitted30 → 60
Success rate (held)13%
Expected funding€4.88M → €9.75M
Cheaper
Per proposal€38k → €5.3k
Annual proposal cost€1.13M → €317k
+ Annual licence€30k
Net annual uplift
Total this year+€5.66M
Funding gain+€4.88M
Operations (incl. licence)+€781k

Caveat: Rough estimate (±10%). The €30k licence figure is the floor — Enterprise pricing is scoped to org size, integration scope, and SSO/audit complexity. The discovery call gives you a firm number.

Our commitments

4 guarantees. In the contract.

Data Sovereignty Guarantee

Your data-flow map is defined before launch and written into the agreement. Enterprise Lite keeps application data in your environment and routes inference only to approved endpoints. Campus Compute keeps both application data and model inference inside your perimeter.

Model policy Guarantee

For Campus Compute, model inference runs on approved open-weight models. For Enterprise Lite, model inference routes only to approved endpoints. We do not change model providers or routing policy without approval.

Exit Guarantee

On termination, you receive the agreed export package: database export, prompts, scorers, templates, workflow configuration, and deployment artifacts. For Campus Compute, local model-weight handling is scoped according to model licences and your agreement.

100% Satisfaction Guarantee

We iterate until your team is producing submission-ready drafts — no cap on rounds, no clock. If we can't get you there under our Fair-Use Policy, the unused portion of the licence is refunded.

FAQ

What IT and procurement ask before signing.

How does the deployment actually happen? +

Enterprise Lite usually takes 3–5 weeks. Campus Compute usually takes 4–8+ weeks, depending on graphics processing unit access, network routing, scheduler constraints, identity integration, and model approval. The first step is a deployment-shape workshop with your information technology and grants teams.

Which models do you use? Can we override? +

For Campus Compute, we select and pin open-weight models during scoping based on quality, language needs, context length, available graphics processing units, and institutional policy. For Enterprise Lite, we route to approved endpoints defined in the agreement.

How is data residency handled if we're outside the EU? +

The deployment lives wherever your IT chooses — including non-EU regions. We have customers in Switzerland, the UK and Norway running entirely on national-cloud infrastructure. The DPA reflects whichever jurisdiction applies; we are ISO/IEC 27001:2022 certified regardless.

What happens if a model becomes outdated or insecure? +

We notify you in writing, propose a replacement from our approved model list, and your IT signs off before we deploy it. For Campus Compute, the old open-weight model stays available in your environment until you choose to retire it. No silent swaps. No deprecation drama mid-proposal.

Can we run completely air-gapped? +

Yes, under Enterprise Campus Compute when the required local compute is available. Fully local inference requires graphics processing units sized for the selected models. Updates can be delivered as offline bundles where required.

What's the support model after rollout? +

Shared Slack channel with the founding team for both your IT and your grants office — no tickets, no support tiers, no AI bots. Quarterly tuning sessions in year 1 to keep prompts and scorers sharp. Custom SLA available for organisations with formal uptime requirements.

How is Enterprise priced? +

Enterprise starts from €30k / year + €5k setup. Enterprise Lite is the faster and lower-complexity path. Campus Compute is scoped separately because local inference, graphics processing unit sizing, network routing, and air-gap requirements vary. No public price list — we'll quote on the discovery call.

Book the call

Pick a slot. Talk to the team.

Usually a CISO or founder on our side. Mutual NDA before any technical detail. 30 minutes, no sales pressure.

Talk to the team that builds it. 30 minutes.

Bring your information technology lead and grants director. We will map the right path: AI Enablement, Enterprise Lite, or Enterprise Campus Compute. No sales reps, no follow-up emails — just the founding team.

Book the discovery call →
Mutual NDA before any technical detail Deployment proposal within 7 days of the call Signed agreement & week-1 kickoff within 2–3 weeks