Governance is not a policy. It is a design practice.
This site brings together four offerings that share a simple belief:
good governance is not about adding paperwork. It is about designing
systems that work before you build them, and proving they existed
when you said they did.
A note before you explore: DBP started as a training framework.
It has since grown into a deployable architecture that powers real applications
through Cambium Engine. You can still learn it in a workshop. But now you can also
run it in production.
Design Before Prompt v2.0 — Dual-Resolution Architecture
Stop governing by gut. Start governing by design.
You have governance frameworks (NIST, ISO, EU AI Act). You have AI tools
(ChatGPT, Claude, Copilot). But you don’t have a method for the work
in between — the reports, decisions, emails, analyses, and systems your
team produces with AI every single day.
Most organizations treat AI governance as a policy document that sits in a drawer.
DBP treats it as a design practice that sits in your workflow.
Tier 1: DBPR
Strategic Cognition Layer
For rapid human alignment, non-technical manager training, and
early-stage scoping — with programmatic hard limits from the first moment.
Start light. Scale governed. Your business-friendly prototype can grow into a
fully audited production system without reinventing the process — and without
leaving security to subjective human judgment. DBP imposes no dependency on
any specific AI model, cloud provider, or platform.
01
Define
What problem are we solving, and for whom?
02
Bound
What should this never do, and who is accountable?
03
Prototype
What does human-plus-AI actually look like?
04
Review
Is this design safe, compliant, and realistic?
Workshops
Format
Duration
Investment
What You Leave With
Foundations
1 day
$8,000
Completed Problem Definition Canvas and Boundary Map for a real project
Implementation
2 days
$15,000–$20,000
Everything above, plus Design Blueprint and Review Checklist, ready for sign-off
Enterprise
4 days
$35,000–$50,000
Full certification, corporate license, and train-the-trainer capability
How DBP fits with existing standards: NIST AI RMF — DBP operationalizes it.
ISO 42001 — DBP sits beneath the management system.
EU AI Act — DBP maps to Articles 6–26.
Free Resources
YouTube: Design Before PromptWhite papers: The Bridge LayerDBP + EU AI ActDBP + NIST AI RMFProblem Definition CanvasBoundary MapDesign Blueprint
Cambium Engine takes DBP’s design principles and embeds them directly
into running applications. It is a serverless middleware that gives your tools
structural access controls, audit trails with cryptographic integrity,
field-level data policies, and continuous validation — all without locking
you into any specific AI model or cloud provider.
Design methodology is useful. But at some point, you need to run it.
Cambium Engine is that bridge.
$0
Infrastructure cost per month (serverless free tier)
Sustainability leadership. Without the full-time hire.
Sustainability is no longer optional. Regulators, investors, and customers
demand credible strategy, transparent reporting, and measurable progress.
But hiring a full-time Chief Sustainability Officer is expensive — and you
may not need one yet.
I work with your team on a flexible, part-time basis to provide strategic
sustainability leadership, framework design, and compliance roadmaps.
📈 Strategy & Roadmap
Sustainability strategy design and development roadmap
📜 Framework Implementation
ESG framework selection and implementation (GRI, ISSB, TCFD, CSRD, ESRS)
📊 Compliance Readiness
Regulatory compliance assessment and gap analysis
👥 Board Advisory
Board-level advisory on sustainability governance and reporting
Best suited for: Mid-size organizations that need expertise without
committing to a full-time executive. The same constraint-first approach applies:
sustainability is not a checklist — it is a system.
Data changes. ETIP ensures no change happens without evidence.
Both AI governance and sustainability reporting share a common challenge:
data evolves. Baselines are recalculated. Reports are revised. Errors are corrected.
The question is always the same: Can you prove what was said, when it was said,
and whether it changed later?
ETIP answers that question. It is a lightweight, post-blockchain protocol for
preserving the temporal integrity of records — without blockchain, without tokens,
without heavy infrastructure.
✓ What ETIP Does
Proves record existence at a point in time
Detects any alteration of past records
Preserves sequence and timeline of changes
Audit-ready evidence, no consensus overhead
✗ What ETIP Does Not Do
Does not verify data correctness
Does not certify regulatory compliance
Does not adjudicate truth or falsity
Does not ensure completeness
ETIP only preserves temporal integrity — the order, existence,
and immutability of records. That is both its power and its limit.
How it supports the other offers: For DBP clients, ETIP can witness
design artifacts. For Cambium Engine, ETIP is the audit trail layer. For sustainability
clients, ETIP preserves every baseline and revision.
Governance is not a policy. It is a design practice. Whether you are designing
an AI-assisted report, a sustainability roadmap, or an integrity layer for
your records — the principles are the same.
Define the problem. Bound the risk. Design the workflow. Prove it is fit for purpose.
About
Yiping Chen
Add your photo here
I work at the intersection of governance, compliance, and operational design.
I created Design Before Prompt (DBP) v2.0 — a design methodology
that helps teams produce better, safer work with AI.
I built Cambium Engine — a middleware that embeds DBP + ETIP
into running applications, making governance deployable without vendor lock-in.
I serve as a Fractional Chief Sustainability Officer — providing
sustainability strategy and ESG implementation for organizations that need
expertise without a full-time executive hire.
I also created ETIP — a lightweight protocol for preserving the
temporal integrity of records. Hosted at
epistemic.center.
My background includes EHS, WSH, RBA, and sustainability framework implementation.
