About · Method

How Ballast-AI engagements run

Assessment → pilot → enablement → optional retainer. Artifacts you can audit. Criteria you can kill.

  1. 01

    Intake

    One metric hypothesis, systems list, risk flags (PHI, OT, procurement). Confidential by default; NDA on request.

  2. 02

    Assessment (2–4 weeks)

    Workflow map, baseline plan, risk/control notes, ranked opportunities, pilot SOW or clear no-go with kill criteria draft.

  3. 03

    Pilot (60–90 days)

    Week-one baseline, thin production slice, weekly readouts, control checks, adoption tracking against the SOW metric.

  4. 04

    Handoff

    Runbooks, ownership, IP clarity, hypercare window. Something your team can maintain on Monday—not a black box.

  5. 05

    Retainer (optional)

    Quarterly ROI, iteration backlog, monitoring. Steady external lead when the first win sticks.

Kill criteria are not a slogan

Paid pilots include explicit go/no-go language. If the baseline doesn’t move—or controls can’t clear risk—we stop or re-scope. You always know what “good” looks like.

Book assessment

Start with one metric—not a 90-slide roadmap

Tell us where cost, throughput, or labor is stuck. We’ll tell you if a 2–4 week assessment is the right next step—or if it isn’t.