Scan & Go

Our implementation model for companies with bigger challenges.

Scan & Go is how h|ve gets deployed. Every phase has an objective, an exit criterion and named deliverables. Nothing moves forward because the calendar says so: it moves when the gate passes.

The first five phases add up to 14 working days. Scaling and evolving follow your company pace.

The seven phases

Every phase ships something, and only ends at the gate.

The shape is always the same: understand where you are, decide how you will operate, build the smallest thing that works, prove it with real squads and then scale. What changes from company to company is the content, not the order.

  1. Kickoff with leadership, technology and business. Why h|ve, which problems to attack and which results to expect.

    Exit gate

    There is a measurable business outcome, a named sponsor, an agreed initial scope and a baseline of the key indicators.

    Deliverables
    • AI Development North Star
    • Business Outcomes
    • KPI baseline
    • Initial scope
    • Candidate squads
    • Stakeholder Map
    • Success Criteria
    Duration1day

  2. Understand the development flow you have today and find where AI removes friction, lowers cognitive load and buys speed or quality.

    Exit gate

    End-to-end flow mapped and an opportunity backlog prioritised by impact, effort and risk.

    Deliverables
    • Current-State Development Map
    • Friction Map
    • AI Opportunity Map
    • Opportunity heatmap
    • Priorization Matrix
    • Baseline per SDLC stage
    Duration3days

  3. Define how people, agents, tools, data and controls will operate together in the target model.

    Exit gate

    Target Operating Model approved, with clear responsibility across human, AI and system, architecture defined and use cases selected.

    Deliverables
    • AI Development Operating Model
    • Target Architecture
    • Human × AI Responsibility Matrix
    • Agent Catalog
    • Target Workflows
    • Guardrails
    • Harness roadmap
    Duration3days

  4. Build and integrate the Minimum Viable Harness, connecting AI to the real engineering environment.

    Exit gate

    The harness works end to end in a controlled environment, with integrations, context, agents and the minimum controls to start the pilot.

    Deliverables
    • Minimum Viable Harness (MVH)
    • IDE, Git, CI/CD and Jira integrations
    • Context Layer
    • Agent Workflows
    • Prompt and instruction library
    • Policies
    • Initial observability
    • Developer Playbook v1
    Duration2days

  5. Check with real squads whether the harness produces measurable gains in productivity, quality and business value.

    Exit gate

    The pilot shows proven benefit without degrading quality, security or compliance, with evidence to decide on scale.

    Deliverables
    • AI Value Proof
    • Before and after metrics
    • Pilot Report
    • Adoption metrics
    • Quality metrics
    • Business Impact Assessment
    • Lessons Learned
    • Business case for scale
    Duration5days

  6. Take the harness from pilot to organisational capability, adopted in waves through a model that repeats.

    Exit gate

    Repeatable rollout, enabled teams, a stable platform and sustainable adoption across several squads.

    Deliverables
    • Enterprise Rollout Plan
    • Adoption Playbook
    • Wave Plan
    • Training and enablement program
    • Champions Network
    • Reusable agent and workflow catalog
    • Support model
    • KPI Dashboard
    Durationto be planned

  7. Turn the harness into a living engineering system, tuned by telemetry, by new models and by what the business needs.

    Exit gate

    Permanent ownership, active governance, a continuous improvement cycle and evolution decisions taken on usage, quality, cost and impact data.

    Deliverables
    • AI Engineering Control Tower
    • Governance Model
    • Model and agent lifecycle
    • Evaluation Framework
    • Cost and usage dashboard
    • Continuous Improvement Backlog
    • Quarterly evolution roadmap
    Durationcontinuous