CONSULTING / APPLIED WORK

Consulting for difficult systems.

Systems & Architecture Review. AI & Decision Governance Review. Knowledge & Research Infrastructure Review.

Bring a system, workflow, decision, or failure that is expensive to misunderstand. We start with the problem you need solved, leave behind useful maps, decisions, prototypes, or operating artifacts, and introduce deeper technical machinery only when the work actually requires it.

SERVICES

Three ways to start, each with concrete first steps.

Choose the problem first. The cards show practical ways an engagement can begin.

Systems & Architecture Review

For consequential systems that are difficult to understand, modernize, integrate, migrate, or repair because their operative model is distributed across code, data, interfaces, workflow, permissions, and institutional memory.

System reconstruction

Reconstruct the minimum continuation-sufficient model: state, interfaces, dependencies, authority, boundaries, invariants, projections, and failure paths required for the bounded change.

Modernization / migration review

Compare repair and migration paths against the behavior, constraints, and semantic obligations that must survive the change.

Bounded implementation / pilot

Turn one selected repair into an executable model, prototype, or pilot with explicit assumptions, authority, witness conditions, failure paths, and handoff state.

AI & Decision Governance Review

For systems where people, software, AI, policy, and automation interact but decision rights, review, escalation, or accountable authority are unclear.

Decision + responsibility map

Trace who or what recommends, ranks, approves, denies, escalates, acts, verifies, and promotes—and where capability and permission have drifted apart.

Governance + approval controls

Define Forge / Certify / Forbid boundaries, human gates, verifier requirements, contestability, provenance, and repair paths for consequential actions.

Decision-chain / failure review

Reconstruct a consequential outcome across people, software, policy, and evidence to locate authority gaps and the smallest credible repair.

Knowledge & Research Infrastructure Review

For organizations that need durable provenance, state, evidence, handoff, and authority across research, documents, schemas, software, and AI transformations.

Source + evidence architecture

Map how source material becomes claims, decisions, artifacts, and promoted institutional state—and where provenance or authority is currently lost.

Information + handoff analysis

Identify where meaning is lost, important views hide assumptions, handoffs require reconstruction, or documents, schemas, interfaces, reports, and teams no longer agree.

Knowledge infrastructure pilot

Design a bounded provenance, research-operations, or shared-knowledge pilot using the smallest useful combination of workflow and machinery.

SYNTHETIC REVIEW EXAMPLE

See exactly what a Systems / Architecture Review is doing.

Start with the represented state, reconstruct the hidden lifecycle, then follow the defect to a bounded repair.

Select to inspect the review

WHAT YOU SHOULD GET

The work should leave behind artifacts, not just conversation.

The exact package depends on the engagement, but the result should make the problem clearer, the decision easier, or the system more operable after BFL is gone.

Select to inspect the deliverables

HOW AN ENGAGEMENT STARTS

Start with the smallest piece of work that can improve the next decision.

Pricing and duration make more sense after the problem and expected deliverable are clear.

Bring a real problem

Start with the system, decision, failure, or opportunity—not a request for a fashionable methodology or predetermined implementation mechanism.

Inspect what exists

Review the software, workflow, documents, data, people, constraints, authority, prior attempts, and the representations through which the system currently knows itself.

Bound the engagement

Respect actual capacity and agree on the smallest coherent review, workshop, prototype, pilot, or advisory scope that can finish, produce evidence, and change the next decision.

Make the work inspectable

Deliver maps, artifacts, findings, code, decisions, or evidence that someone other than the consultant can examine.

Inspect and adapt

Treat the result as new system state: repair, build, continue, transfer, pause, or stop based on what the engagement actually revealed.

CONSULTING PRINCIPLES

Useful consulting should reduce ambiguity without manufacturing certainty.

Boundary First Labs is most useful when the engagement can make state, responsibility, assumptions, evidence, and repair more explicit.

LEAN–AGILE BY PRACTICE, NOT CEREMONY

Use visible state, bounded work, real capacity, small coherent increments, and short evidence loops. Test important assumptions instead of requiring clients to adopt a process brand.

HUMAN AUTHORITY STAYS VISIBLE

AI may search, compare, simulate, draft, or critique, but consequential decisions remain with the people and domain experts who actually hold the relevant authority.

STEWARDSHIP OUTLIVES DELIVERY

Account for maintenance, repair, transfer, retirement, affected people, and material consequences. Useful work should remain operable after the engagement ends.

NO THEORY BUY-IN REQUIRED

Clients do not need to learn Boundary First terminology or agree with the Lab's research program. The engagement has to stand on the usefulness of the work itself.

SCOPE BEFORE SCALE

Prefer a review, workshop, prototype, or pilot that can finish and create evidence before proposing a large transformation.

EVIDENCE OVER PERFORMANCE

Findings should expose assumptions, uncertainty, disagreement, and the evidence needed for the next decision—not merely produce a confident-looking deck.

A GOOD FIRST ENGAGEMENT

Bring one system that is expensive to misunderstand.

A useful first step may be a review, workshop, diagnostic, prototype, or bounded pilot. If the work creates evidence and the next problem becomes clearer, the relationship can grow from there.