Request a six-week shadow evaluation.

Build the case for governing one consequential workflow. Compare FieldHash with your existing controls, with production enforcement off.

Start with three details.

  • What consequential action or decision does the workflow make?
  • Which policy, approval, or reviewed decision authorizes it?
  • What can change before that authority is used?
Email your workflow

Record what FieldHash would allow, withhold, or send to review. Production enforcement remains off.

You do not need a finished integration plan. We will reply with fit, scope, and next steps.

What you walk away with

Traces of what would have governed each handoff, review burden and latency findings, and an evidence report your legal, security, and governance teams can inspect. Where relevant, controlled replay shows which stale records or revoked actions FieldHash excluded.

Agree the scope and acceptance criteria before evaluation. At closeout, decide whether to enforce, adjust, continue in shadow mode, or stop.

Reviewers investigate shadow disagreements; these are not prevented incidents. Before production enforcement, separately test how the agent responds to a denial and complete the acceptance review.

Start where your team keeps asking: does that approval still apply?

Bring a live or near-live workflow, an accountable owner, and a reviewer who can establish what should govern. A policy assistant with changing exceptions or an agent with several routes to an approval-gated action makes a useful starting point.

Evaluate changing approvals, reviewed-decision reuse, instrumented action paths, or per-workflow cumulative limits where configured.

Current pilot scope: The shipped kernel supports per-workflow cumulative budgets. The shared-budget runtime evaluated across cooperating agents is a qualified research candidate, not part of the current pilot deployment. Cross-agent shared-budget requirements need separate scoping.

Pilot structure

What's included in the 6-week pilot

The single-tenant reference deployment provides an authenticated gate, private operator console, durable review queue, signed ledger, and backup-and-restore tooling. The SOW records which components the pilot uses, their access controls, and the recovery checks required before acceptance.

One scoped AI agent, RAG workflow, enterprise memory source, tool workflow, MCP stdio or stateless HTTP server, or instrumented computer-use path
Baseline capture of the current retrieval, memory, or action handoff
Shadow-mode comparison showing what would proceed, stop, or go to review
Controlled enforcement simulation showing excluded records or actions removed from the packet
Governed Precedent review where prior reviewer decisions may safely carry forward
Governed Learning lineage from the initial packet through review, Authority State, reuse, suspension, and return to review
Source-health and freshness report for connected authority sources where configured
Change-impact report showing affected workflows, precedents, and review routes
Reviewer-ready evidence packet, scorecard, closeout findings, and verification commands where configured
A production-scope recommendation: enforce, revise the authority inputs, continue in shadow mode, or stop

Send the workflow. We will reply with fit, scope, and next steps.

Tell us where an approval, policy, or earlier review needs to keep governing as the work changes. We will agree the integration scope, data access, operating constraints, and measures of success with your team.

Optional details for scoping

Share these if you already know them. They are not required to start a conversation.

  • Where approved, revoked, or superseded state lives, and who reviews it.
  • One approved, expired, or rejected example relevant to the workflow.
  • An available authority source, such as prepared metadata, a read-only SQL view, or a signed event feed.
  • The agent or tool path, and any model or deployment constraints.

Need the package first? Review pilot scope. Need the architecture? See how FieldHash fits into your stack. Need the packet shape? See the product and deployment components.

pilots@fieldhash.ai

Typical reply within 2 business days.

We use what you send to respond, scope the evaluation, and support requested deployment review. We do not sell personal data. See our Privacy Policy.