Platform
The AI proposes. Your policy decides.
Seventeen specialist agents are embedded across the application, doing the day-to-day work beside you. Every one of them runs under the same policy, inside your own signed-in session, and records every exchange it has — because in a system whose output has to survive an audit, what matters is not what a model can do but what it is allowed to do, and whether you can prove what it did.
The financial problem
Automation you cannot explain is automation you cannot deploy.
- An accrual adjusted by a model, with no record of why, is an audit finding waiting to happen.
- An assistant with an account of its own is a back door: anyone who can type a question reads at its level.
- Time saved is worthless if every output has to be re-checked manually anyway.
- The value is in the volume work: matching, mapping and anomaly detection at a scale humans cannot cover.
The technical problem
Scope, policy and trace are the three controls that matter.
- Scope: the agent has no account or access of its own — it works inside the asking user’s session and company workspace.
- Policy: what agents may do is configured at product, company and role level, not hard-coded.
- Guardrails: Ask AI requests and every reply are checked; anything flagged routes to a managed review queue.
- Trace: every exchange — and exactly what was sent to the model — is recorded, encrypted and explainable after the fact.
How it is governed
The RevUpra Agent, and every specialist under it.
Agents are not a collection of separate tools that each need their own controls. One agent stands in front of all of them, and every other agent answers to it — which is what makes a single policy enforceable across the whole estate.
The RevUpra Agent
One front door
Every Ask AI request enters through the RevUpra Agent. It decides which specialist should handle the work, passes the caller's identity along with it, and owns the answer that comes back. In-page assistants go straight to their specialist — under the same policy layer.
Always-on guardrails
Checked in both directions
A request filter screens every Ask AI question before a model sees it, and a response filter checks every answer before a person does — removing internal detail and masking the fields your company marks as sensitive. They are agents too: configured and audited like the rest.
Skill agents
Narrow on purpose
Agreement builder and analyst, rebate analyst, query assistant, field mapper, contract reader and reviewer, deal efficiency, market scout, billing reconciliation, task creator, app builder and the rest. Each does one job, so what it can ever touch is limited to that job — a general-purpose agent has to be trusted with everything.
Policy
Three levels, merged key by key
Policy is configuration, not code. Each level narrows the one above it, and they merge field by field — so a role can tighten a single threshold without restating a policy it otherwise agrees with.
-
01
Product
The baseline every company starts from — what agents may read, what they may propose, and what they may never do. Maintained by RevUpra and applied to every call.
-
02
Company
Your controls environment. Turn a capability off entirely, lower a threshold, or require review where we would have allowed a direct action.
-
03
Role
The narrowest level. A role-scoped policy applies only to users holding that role, so an analyst and a finance approver can face different limits on the same agent.
An agent with no applicable policy does not get a default of "allowed". It gets the product baseline, which is the most restrictive of the three.
Access
No access of its own. It borrows yours.
An agent is not a service account. It runs as the person who asked.
Your session, your company
Everything an assistant reads, it reads through your own signed-in session, inside your company's walled workspace — the same database-level isolation as every screen. There is no elevated identity behind the assistant, because an identity with god rights is the single thing that makes every other control cosmetic.
It can only create what you could
Assistants read, explain and propose. The few things one may create — an agreement drafted from a document, a follow-up task — are created as you, only if your role allows it, and as a draft that goes through your normal approval and signature steps.
Audit
Every token is accounted for
Not a sample. Not an aggregate. Every interaction, with what it cost and what it touched.
Token spend, per call
Tokens in and tokens out are recorded against the agent, the user and the page it was asked from — so spend is attributable, not a monthly surprise.
Flagged, not buried
Anything a guardrail catches is raised to system administrators with the prompt, the response and the rule that fired.
Reviewable after the fact
An administrator can open any exchange and walk its call chain — what was asked, which agent answered, on which engine, and what came back.
Exactly what was sent
The record keeps what left for the model in its parts — the data it was grounded on, the history, the question — so an answer can be traced to its inputs.
Bring your own agent
Connect any agent. Including the one you already run.
The provider is configuration. Point the RevUpra Agent at our default, at your own enterprise agent behind your own contract and data-residency terms, or at whichever model your security team has already approved. The governance layer does not move: whatever is connected still runs under the same policy, still works inside the caller's session, and still records every exchange.
That is the point of governing the seam rather than the model — you can change your mind about the model without renegotiating your controls.
Context
It already knows what you are looking at.
Most questions are about the screen in front of you. So Ask AI starts there — and never has to ask which agreement you mean.
The screen, the record, the tab
Open an agreement on its Lines tab and ask “why isn’t this line earning?” — the assistant already has the agreement, its allowances, tiers, lines and qualifiers.
Read live, with your access
Your browser sends only which record you are on. RevUpra reads it itself, through your own session, and masks sensitive fields before anything reaches a model.
One tab at a time
With several RevUpra tabs open, the assistant sees only the one you asked from — and says so, rather than guessing about another tab.
Yours to switch off
A chip shows what the assistant is looking at. One click leaves the screen out of a question.
The right engine for each job — and what reaches it
Each specialist can run on its own connection, so quick helpers can use a fast model while the ones doing analysis use a stronger one. Provider keys are held in an encrypted vault and are never shown on a screen.
Each question is sent as one self-contained request carrying only what that task needs. RevUpra keeps no files, trained model or index at the provider; how long a provider keeps its own request logs follows that provider's terms — choose a connection under your enterprise agreement for the strictest posture.
What it does
Agents doing the volume work
The tasks worth automating are the ones that are too numerous for people and too rule-bound to be interesting.
In-app assistant
Answers questions about the agreement, claim or contract on your screen — its terms, tiers, lines and qualifiers — from the live record.
Contract capture
Reads an executed agreement and proposes the structured terms for a human to confirm.
Field mapping
Proposes integration mappings for a new partner feed from the shape of the data.
Why didn’t this earn?
The Rebate Analyst explains a transaction that missed a rebate, from a check that has already run — never a number it made up.
Query assistant
Writes a read-only query over the tables your administrator approved. It cannot run it; you do, with the usual limits and masking.
Review queue
Everything flagged, in one place, with the proposal, the reasoning and the accept or reject decision recorded.
The line we draw
There is a real distinction between AI that answers and AI that acts. An assistant that reads inside your own session is low-risk and immediately useful. An agent that adjusts an accrual is a different thing entirely, and needs an explicit policy, a review gate and a record of what it did and on what basis. Today RevUpra's assistants answer and propose; changing a record is a person's job.
As agents that act arrive, the boundary between answering and acting stays yours to draw — because where it belongs depends on your controls environment, not on ours.
Ask the assistant on this site.
If it is switched on, the chat button in the corner is the same governed-assistant pattern, pointed at public content instead of your ledger. Ask it something hard.