PRACTICE FRAMEWORK — HOW ENGAGEMENTS RUN

The evidence-chain
engagement methodology

Consulting engagements usually end with a slide deck and a vague sense that value was delivered. This methodology ends differently: every phase terminates in named, versioned, attributable work product — so what you paid for is a matter of record, not recollection. The engagement is run on the same evidence discipline it typically architects for the client.

5 PHASES
100+ ARTIFACTS PER ENGAGEMENT
1 DELIVERABLES MANIFEST

Phases overlap chronologically — they are distinguished by intent, not calendar boundaries.

PHASE 1 — IMMERSE

Arrive knowing more about your market than you expect

Before the first standup, the engagement opens with a pre-start intelligence sprint: market sizing, competitive analysis, and a strategic thesis grounding your platform's position in structural terms. Board-grade material most organizations lack internally — delivered before formal work is assigned, so the technical work that follows is already anchored to your commercial reality.

REPRESENTATIVE ARTIFACTS
Domain research paper / market thesis Multi-year executive strategy Market forecast (TAM / SAM / SOM) Competitive win/loss framework Annual blueprint + quarterly OKRs Board deck Target-state solution architecture Current-state architecture (C4 / Mermaid) Capability map (L0–L2) Operating model canvas Ecosystem & stakeholder map Capability heat map Vendor landscape assessment Regulatory environment brief M&A / exit positioning memo
PHASE 2 — DIAGNOSE

Evidence before prescription

No recommendation is made without a documented, quantified deficiency behind it. The diagnostic phase produces the hard numbers — remediation effort versus available capacity, integration revenue at exposure, incident cost — so that every architectural decision later in the engagement traces back to a finding you can point to.

REPRESENTATIVE ARTIFACTS
Platform stabilization assessment Technical debt assessment (quantified) Infrastructure architecture + resource inventory Integration catalog + stability assessment API test suites Build / Buy / Blend analysis Engineering standards Security & vulnerability assessment Governance gap analysis Data quality assessment Identity resolution analysis Threat model DR / business continuity gap review Cloud cost / FinOps assessment SDLC / DevOps maturity assessment Incident root-cause analyses
PHASE 3 — ADAPT

Redirection without waste

Priorities change mid-engagement — that's normal, not a failure. When leadership redirects, output pivots on the directive while earlier work stays in active circulation; nothing produced gets orphaned. The pivot itself is documented: the directive, the response, and the impact on scope all enter the record, protecting both parties.

REPRESENTATIVE ARTIFACTS
Operating cadence & standing meetings Dated status snapshots Stakeholder-specific architecture views Team / POD swimlane structure Priority framework RACI matrix Architecture Decision Records (ADRs) Scope-change memo with impact assessment Roadmap re-baseline
PHASE 4 — CONVERGE

Parallel workstreams, one canonical core

The build phase runs multiple workstreams simultaneously — but they converge on shared anchor artifacts rather than diverging into silos. The canonical data model is the gravitational center; everything references it. Traceability is the connective tissue: every rule maps to its regulatory citation, its data entity, and its exception workflow. R&D spikes test vendors against your live documents before any build-versus-buy commitment.

DOMAIN RULES / COMPLIANCE
  • Declarative rules framework (DRL / JDM)
  • Rule traceability matrix
  • Rules engine architecture
  • Risk scorecard PRD + model
  • Document control framework
PLATFORM / DATA ARCHITECTURE
  • Canonical data model (bitemporal DDL/DML)
  • Versioned domain model
  • Ingestion & integration architecture
  • Migration scripts + summaries
  • POC solution architecture
PRODUCT & INTEGRATION
  • Product PRDs + ingestion specs
  • OpenAPI 3.x client specifications
  • IaC enterprise scaffold (Terraform)
  • Reporting implementation guides
  • Integration trackers
R&D SPIKES
  • Extraction tests on live documents
  • Classifier research (zero-shot / trained)
  • Extraction schema definitions
  • Vendor deep-dives
  • Reference implementations
Frequently added: Schema registry + versioning policy Data governance policy NFR matrix Observability specification Capacity plan ML model governance framework Data retention & archival policy
PHASE 5 — TRANSFER

A clean handoff and a documented record

The engagement ends with your team building on the artifacts, not depending on the consultant. Successors onboard against the produced work in the final weeks — the canonical model becomes your infrastructure the moment others build on it. The terminal artifact is a deliverables manifest: every artifact dated, classified, and valued, with standalone value attributed against contract cost. What was delivered is a matter of record.

REPRESENTATIVE ARTIFACTS
Deliverables manifest (dated, classified, valued) Forward plan / SOW extension with ROI attribution Operational findings summary Process guides for inherited workflows Final revision of the domain research paper Runbooks (operational, incident, recovery) Onboarding curriculum per successor role Wiki structure + seeded content Recorded artifact walkthroughs Open-items register with owners
WHY IT WORKS

Operating principles

Evidence-chain reflexivity

The engagement's own work product meets the same standard it prescribes: immutable, provenanced, complete, standardized, reconstructable.

Diagnosis precedes prescription

Quantified deficiency documentation anchors every recommendation. If it can't be traced to a finding, it isn't proposed.

Spikes over speculation

Build-versus-buy decisions rest on empirical tests against your live data — never on vendor claims or analyst quadrants.

No orphaned artifacts

Re-scoping redirects future output; it never strands prior output. Everything produced stays in circulation and on the record.

Canonical gravity

One anchor artifact — typically the canonical data model — that all workstreams reference. Convergence is designed, not hoped for.

Exit as value event

Knowledge transfer converts deliverables into your infrastructure; the manifest makes the delivered value auditable long after the engagement ends.

See it applied to your platform

Recent application: a compliance platform engagement producing 115 artifacts across 52 working days — strategy, canonical data model, declarative rules framework, and full knowledge transfer.