How we work with enterprises

Start with proof.
Scale with confidence.

Large organizations buy evidence before scale. We align on metrics, run a bounded pilot on your data, then expand module by module or as a governed suite in your cloud.

Indicative
pricing bands.

Indicative starting points in USD, excluding taxes. Your written proposal reflects document volume, agent count, deployment model and connectors after the demo.

  1. Recovery audit

    Quoted after sample

    Run a free sample on your own export. The full audit is a fixed fee or a share of confirmed recoveries, quoted once you have seen the sample results.

  2. Paid pilot

    From $15,000

    4–6 weeks

    One workflow with written success metrics. The pilot fee is credited against the first year if you continue.

  3. Agent module

    From $40,000 a year

    per year

    One ProcureAI or FinanceAI agent in production, sized to your document volume and review model.

  4. Suite

    From $120,000 a year

    per year

    All agents in ProcureAI or FinanceAI, with evaluation and support aligned to your governance needs.

  5. In your cloud

    Custom quote

    SSO, custom connectors and a dedicated environment in customer-managed cloud — scoped after security and integration review.

  1. Demo

    Your documents or ours

    See the agents work on procurement or finance documents so your team can judge fit before anyone shares production data.

    • Live walkthrough of coordinated agent roles on sample or redacted files
    • Questions on your workflows, systems and governance expectations
    • A clear view of what a paid pilot would need to prove
  2. Annual subscription

    Per module or full suite

    Keep proven agents running in production: one module at a time or the full coordination layer, sized to your document volume and review model.

    • Agent modules aligned to intake, contracts, invoices, sourcing or spend
    • Evaluation and drift checks tied to your agreed metrics
    • Model and provider updates as your stack evolves
    • Priority support during business hours in your region
  3. Enterprise

    Your cloud · your identity

    Deploy the suite where your security team already governs data: your cloud accounts, SSO, audit trails and contracted support levels.

    • Deployment in customer-managed cloud with your identity provider
    • SSO, role-based access and audit logging for reviews
    • Agreed support tiers and escalation paths
    • Implementation and integrations scoped as a separate fixed quote

Agent products
from Stellitron.

Packaged agents for specific domains, built on the same orchestration and review patterns we use in custom engineering work.

ProcureAI by Stellitron

Live

Coordinated agents for procurement intake, contracts, invoices, suppliers, sourcing, spend and recovery audit — with a demo workspace and trial on your own sample documents.

See ProcureAI

FinanceAI by Stellitron

Live

Agents for bank reconciliation, KYC onboarding and month-end close, each with a visible decision trace — with a demo workspace and trial on your own sample documents.

See FinanceAI

HealthAI by Stellitron

Coming soon

Administrative and operations workflows for health systems, with the same human-in-the-loop pattern: agents propose, your staff approve.

Request early access

Custom
engagements.

When you need agents outside a product module — or deeper integration with ERP, CLM and identity — we scope discovery, pilot and run phases the same way as product rollouts.

  1. Discovery sprint

    2 weeks · fixed scope

    Map one business workflow end to end and agree, in writing, what a pilot must prove before anything is built.

    • Workflow, data and systems map
    • Success metrics agreed in writing before any build
    • Architecture, data-access and security plan
    • Risks, human checkpoints and evaluation approach
    • A fixed-price pilot proposal

    The fee is credited in full against a pilot started within 60 days.

  2. Pilot

    6–8 weeks · fixed scope

    Build one workflow, run it on your data in a controlled environment and measure it against the agreed metrics.

    • A working agent workflow with human approval points
    • An evaluation suite tied to the agreed metrics
    • Weekly working demos with your team
    • A go / no-go review with a production plan and costs

    Paid in three milestones. Stop at any milestone and keep everything built so far.

  3. Run & improve

    Monthly · 30 days’ notice to stop

    Keep live agents accurate, affordable and up to date as models, providers and your business change.

    • Ongoing evaluation and drift checks
    • Model and prompt updates as providers change
    • Cost tuning and a monthly report
    • Priority support
  • You own the code and intellectual property you pay for.
  • Model and cloud usage is billed at cost with no markup, or runs in your own accounts.
  • An NDA is in place before any of your data is shared.
  • Commercial terms are set out in a written proposal once scope, volume and deployment are agreed.

Built for how large
organizations buy.

Enterprise AI pilots often stall between a promising demo and an approved production system. The engagement is shaped to get through that gap.

A written success plan before the pilot.

Metrics, data access and acceptance criteria are agreed with your stakeholders during discovery, so the pilot ends in a clear decision.

A pilot sized for a decision.

Four to eight weeks is long enough to run on real data and short enough to keep momentum through review boards and budget cycles.

Security questions answered early.

Data handling, model providers, access and audit trail are documented in discovery, before your security and AI governance reviews ask.

Your cloud, your control.

Systems can run in your own cloud accounts with your identity provider, so data stays inside the boundaries you already govern.

Common
questions.

How is pricing set?

The bands on this page are indicative starting points in USD, excluding taxes. Your quote depends on document volume, how many agents you run, whether we deploy in our environment or yours, and which connectors you need. After a demo and scope conversation we send a written proposal so procurement can compare a fixed number to your reality.

What if the pilot doesn't meet the agreed metrics?

The go / no-go review says so plainly, with what we learned and what it would take to close the gap. You decide whether to continue; there is no obligation to move to production.

Can you work within our procurement process?

Yes. We can review your vendor agreement, security questionnaire and data processing terms during discovery so contracting doesn't hold up the pilot.

Is the discovery sprint worth it if we already know what we want?

Most AI pilots that stall never agreed what success meant. Discovery fixes that in two weeks, and its fee is credited against the pilot when you start within 60 days.

A GOOD PLACE TO START

Bring the ambition.
We’ll build the way forward.

Let’s talk about your project

A conversation about your goals, constraints, and next step.