Practice 01 · Systems · Sheet S4.1
Decide
before
you build.
Architecture reviews, build-vs-buy analysis, system audits, and technology roadmaps. From people who build for a living, priced so the advice stays honest.
Title block
- Sheet
- S4.1: Advisory
- Practice
- 01 · Systems
- Scope
- Reviews, audits & roadmaps
- Engagement
- Days to weeks, fixed fee
- Deliverable
- A decision you can defend
- Status
- Taking on engagements
Fig. S4.2The symptoms
Expensive decisions, loud opinions.
Technology decisions arrive with pressure from every side: vendors selling, deadlines looming, everyone certain. It usually sounds like this:
A vendor quote is on your desk, and you can’t tell whether it’s fair or fantasy.
Your last build failed, and nobody can explain what actually went wrong.
The board wants “digital transformation” and you have to turn that into a plan.
Everyone on the team has an opinion. Nobody has compared the numbers.
You’re about to commit serious money on the word of the people selling it to you.
Every one of these is a decision, not a project. Deciding well is the cheap part; get it wrong and you pay for it twice.
Fig. S4.3The instruments
Five ways to buy certainty.
Advisory here isn’t a retainer or a workshop series. It’s five fixed-scope instruments, each pointed at one kind of decision.
Architecture review
Your system, read the way an engineer reads it (code, schema, and infrastructure) and returned to you as a marked-up drawing with the risks ranked and the fixes ordered.
- Codebase read
- Risk ranking
- Fix list
Build-vs-buy analysis
The comparison nobody selling to you will do honestly: products versus custom versus hybrid, with five-year costs, and a verdict, not a shrug.
- Five-year math
- Options table
- Verdict
System & vendor audit
The platform you inherited or the proposal you received, independently verified. What’s sound, what’s fragile, what’s overpriced. In writing.
- Code audit
- Proposal review
- Rescue plan
Technology roadmap
Where your operation is going, translated into systems: phased, sequenced, and budgeted in stages a board can actually approve.
- Phases
- Budgets
- Sequence
AI readiness review
What AI would genuinely do for your operation, whether your data can carry it, and, just as valuable, the list of things to skip.
- Use cases
- Data check
- Skip list
Fig. S4.4The docket
Asked, and answered.
A composite docket: the questions owners actually bring us, and the shape of the answers they leave with.
“Our operations run on spreadsheets. Do we need a custom platform?”
A proven product covered 80% of it. Custom could wait until it couldn’t.
“The system a previous vendor built is failing. Rebuild from zero?”
Two of five modules were salvageable. Rescue first, rebuild in phases.
“Should AI be in next year’s budget?”
The data foundation wasn’t ready. Fix that first; then the AI pays.
“We’ve outgrown every tool we’ve tried. Custom now?”
Third time hitting the same wall. The five-year math finally favoured owning it.
“Two vendors, two proposals, zero clarity. Which one is right?”
Both quoted the wrong problem. The scope that fixed it was half the size.
“Can our platform survive ten times the volume?”
Not as built. The audit named three bottlenecks, and the order to fix them in.
Verdicts of “build”1 of 6
Every findingYours, builder-agnostic
One “build” out of six. That’s what advice looks like when it isn’t a sales funnel.
Fig. S4.5A straight answer
Do you even need an advisor?
Advisory is the smallest thing we sell, and we’d still rather you skip it than waste it. The honest split:
Skip the advisory when
- The decision is small enough to just try: a week of building beats a month of deciding
- You’ve already decided and want a rubber stamp: validation theater helps nobody
- The honest blocker is budget, not clarity
- You need a scapegoat more than an answer
A second opinion pays when
- A six-figure commitment is about to be signed
- Two vendors are telling you opposite things with equal confidence
- You’ve inherited a system nobody fully understands
- The board needs an independent view before it votes
And the bias question, answered: fixed fee, findings yours, and most verdicts don’t end in “hire us to build it.”See how we build, when building wins
Fig. S4.6The engagement
From question to verdict.
Short, dense, and evidence-first. We read your actual artifacts (code, contracts, data) before we form a single opinion.
The brief
daysYou bring the decision; we scope which instrument fits it and quote a fixed fee. If the question isn’t worth paying to answer, we say so on this call.
A fixed price on a question worth answering
Fieldwork
1–2 weeksWe read the evidence: codebases, schemas, contracts, invoices, exports. Interviews come second: artifacts don’t have opinions.
Findings built on evidence, not impressions
The verdict
one working sessionThe answer, with numbers, argued to your face, where you can challenge it live. Not a deck couriered over and never opened again.
A decision you can defend upstairs
The paper
daysThe blueprint, roadmap, or audit: written so any competent builder can act on it. It names systems and steps, not buzzwords.
A document that outlives the meeting
Stand behind it
ongoingQuestions surface after the invoice; we stay reachable. And if a build follows, any builder can run with the paper, ours only if you choose.
An answer that doesn’t expire
The house rules.
- Fixed fee, upfront
- Findings builder-agnostic
- We read code, not just slides
- No 80-page decks
Fig. S4.7Asked often
Questions, answered straight.
Won’t you just tell us to hire you to build?
The docket above is the honest answer: most verdicts aren’t “build,” and plenty aren’t “build with us.” Advisory is fixed-fee and the findings are written builder-agnostic: the instrument only works if the pencil stays neutral, and our reputation rides on that.
What do we actually receive?
Depends on the instrument: a marked-up architecture with ranked risks, a build-vs-buy comparison with five-year numbers, an audit with a rescue plan, or a phased roadmap with budgets. Always the same spine: a verdict, the evidence behind it, and a document any builder can act on.
How long does it take, and what does it cost?
Days to a few weeks: these are the smallest engagements we sell, deliberately. The fee is fixed and quoted after the first call, once we know which instrument your question needs.
Can you review a vendor’s proposal without burning the relationship?
Yes. It’s one of the most common requests. The review is confidential, we speak fluent vendor, and the usual outcome isn’t a fight: it’s a sharper scope and a fairer price on both sides.
Our last system failed. Can you tell us what went wrong?
That’s a standard audit. We read what was built, name what’s salvageable, and write the honest post-mortem: no blame theater, just what happened and what to do next. Rescues often cost far less than the rebuild everyone assumes is needed.
Will you sign an NDA?
Of course, standard practice. Your code, contracts, and data stay confidential, and anything we learn advising you is never reused elsewhere.
Cross-references
Read alongside this.
The pages that usually get opened next, and why.
Fig. S4.8Start here
Bring us the decision you’re stuck on.
One conversation to scope it, a fixed fee to answer it, and a verdict you own either way. The most expensive option on the table is guessing.
Malaysia · Working worldwide