PIERRE · COMPASS
PROPOSAL · NEXT PHASE

Answer what we can.
Measure what we can't.

The partners are right that the knowledge base is the real product and the assistant is only its interface. Where the conclusion goes wrong is the sequencing: knowledge first, AI second is an open-ended internal project with no deadline and no measurable output. Run them coupled instead — and let the system produce the backlog.

6
Survey responses
5
Knowledge areas
8
Proposed builds
2 wks
To first backlog
EVIDENCE

Where this came from

Two partner questionnaires — six substantive responses, two Hungarian and four Serbian / Croatian. That is a signal, not a statistic: it tells us what to look at, and the live ticket flow will decide the actual ranking.

What partners reported

Software and firmware faults, and updates that reached the field under-tested. Slow response when a developer has to be involved. No diagnostic tooling, no register of known faults. Installers left without help outside working hours — and repeat site visits that cost money directly and reputation on top.
pain

What they asked for

A support portal, a direct line to an engineer, diagnostic tools, faster response, remote assistance, wiring examples, a searchable register of known faults and firmware versions, and a real L1 → L2 → developer path with committed response times.
wishlist

The two that discriminate

Most of that wishlist any hardware vendor's partners would produce. Two items are specific to PIERRE and worth more than the rest combined: firmware quality, and distributors acting as informal L1 without the access or backing to do it.
signal
THE SHIFT

What the system is for in phase one

Today the goal is: answer well. For the next phase it should be: answer what it can — and manufacture the list of what it can't.

WHY IT MATTERS

“Digitise the company's technical knowledge” is correct and unbounded. No owner, no deadline, no visible progress — the shape of task that quietly stops. “Write these three articles this week, they cost you eleven escalations” is finite, and someone can be held to it.

The instrument is already in the codebase. The confidence engine emits LOW and SAFETY_RISK on every turn it can't support with evidence, and the audit chain records the retrieval that failed. Nobody collects it. Collecting it is most of item D.

THE DELIVERABLE AT WEEK 2

Not a feature. A ranked list of what to write, ordered by how much each gap actually costs in escalations and site visits.

Run it on internal support first, not partners — the internal team tolerates a bad answer and their questions fill the ledger. Partners get it once the top of the list is written.

HONEST BALANCE

Five knowledge areas vs. what exists

The partners named five things to digitise. Of those, one is structurally complete, two are half-built, and two are genuine gaps — plus the metadata contract, which is missing its more important half. Status below is read off the repository, not the spec.

Product & version master
PARTIAL
Product, ProductPin and WiringRule are already mirrored from B2B behind a live connector — and B2B stays the source of truth, so a second product master would be duplicated work. What genuinely doesn't exist is the version axis: no firmware or hardware release entity anywhere. The word firmware appears in the codebase in exactly one role — as a ticket category.
Install & config knowledge
PARTIAL
The pipeline is done and running: the wiki adapter pulls Academy lessons from B2B live, YouTube transcripts are chunked with timestamps, and there's even a connection-diagram generator. The open item is content volume — ~300 wiki pages and ~200 videos are a plan, not a verified ingest.
Troubleshooting procedures
MISSING
There is a diagnostic mode and an intent classifier, but no structured procedure entity. RAG over prose is not a guided procedure. The regression set shows this precisely: on the Berlin-display case the model named the real root cause — app firmware updated, launcher didn't — because a human put that case into golden-examples.ts, not because a documented procedure existed for it to follow.
Known issues & changes
MISSING
No model, no ingestion, no tool. And per the survey this is the single largest source of partner frustration — firmware and software faults, and updates that reached the field under-tested. The biggest gap between what matters most and what exists.
Resolved cases & feedback loop
BUILT
Structurally complete: SupportTicket, PendingTicketReview, a promote-to-KB review queue, trust levels and an append-only audit chain. The loop exists — it just has no historical corpus loaded, and no “seen three times, write an article” trigger.
Ownership metadata
PARTIAL
Present: version, language, trust level, status, lastIndexedAt, freshness SLO panel. Absent: an accountable owner and a next-review date — the two fields that turn a document store into knowledge management.

Read together: the machine is largely finished and the fuel is missing. The larger half of the remaining work sits with PIERRE, not with engineering — which is good news, provided the list of what to write is generated rather than guessed.

PROPOSAL

Eight things to build

Each one names what already exists underneath it, because most of these are smaller than they sound — the platform was built to take them.

A

Version axis

ProductVersion · feature→version matrix · update-risk flags

The foundation most of the rest sit on. A firmware / hardware / app release entity with release notes and a predecessor chain, KnowledgeSource.appliesTo scoped to a product and version range, and version filtering inside retrieval, so an article written for 1.8 stops surfacing on 2.4. Session state grows a case-context slot (product, version, environment) that the assistant fills in up front rather than halfway through.

Two additions from the operational concept turn this from a filter into an answer generator. A feature → minimum-version matrix, so the assistant can say “that feature needs v20872, the device runs v17659” — which is one of the questions partners actually ask. And an update-risk classification per release: routine, or requires on-site presence, or requires PIERRE approval. An update that needs a physical power-cycle if it fails is a different object from a routine one, and nothing records that distinction today.

WHAT'S ALREADY THERE

Session state and the retrieval scoring path already exist. First task is a question, not code: does B2B track firmware releases? If yes we sync them. If not, this is the one place a new master is justified — release history exists nowhere today.

One hard rule ships with it: the assistant never initiates an update and never changes a configuration. Not as prompt guidance — as an absent tool. It shows the required version and the risk; a permitted partner triggers the update themselves in the Portal.

B

Known-issue registry + proactive alerts

KnownIssue · checkKnownIssues · push

Affected products and version range, symptom, how to recognise it, severity, workaround, what not to do, expected fix, fixed-in version, owner, last verified. The tool runs before retrieval on the diagnostic path, because a known issue outranks a wiki article — and it feeds the confidence engine as a structured high-trust signal, the way wiring rules do, not as another RAG chunk.

WHAT'S ALREADY THERE

The part that makes this worth more than a database: when an issue is filed, the system already knows whose buildings run the affected firmware — the Portal has the device inventory and PushSubscription is in the schema. The installer learns about a bad release before standing in front of it, workaround attached. That turns the survey's loudest complaint into the one thing a support portal can never do.

C

Diagnostic procedures as data

DiagnosticProcedure · getDiagnosticProcedure

Symptom → ordered checks → what to observe or measure → branch → resolution or escalation, plus an explicit stop-here condition. Structured rather than prose for two reasons: the model must not improvise the order of electrical checks, and a fixed procedure makes it deterministic which data gets collected — which is what makes the resulting ticket useful.

WHAT'S ALREADY THERE

Authoring is the expensive part, so don't staff it. Generate draft procedures from resolved tickets and have a human edit and approve — exactly the pattern PendingTicketReview already implements, just producing a procedure instead of a Q&A pair. That is how you get thirty procedures in weeks instead of quarters.

D

Knowledge-gap ledger

KnowledgeGap · closure gate · /admin/gaps

A normalised question cluster with first and last seen, hit count, best confidence reached, escalation rate, affected product and version, status and owner. Fed automatically from every LOW and SAFETY_RISK turn, every empty retrieval, every escalation. Clustered by embedding similarity so “the same question forty different ways” is one row.

Paired with it, from the operational concept: a closure gate. A ticket cannot be closed until the actual cause is recorded — including when it wasn't ours, with its own taxonomy: internet outage, severed cable, lightning, physical damage, work by someone unqualified. That second list is a business metric in disguise. It shows what share of “PIERRE faults” were never PIERRE faults.

WHAT'S ALREADY THERE

The cheapest item here and the most important. The data is already produced by the confidence engine and the audit chain — today we throw it away. Ranked by frequency × escalation cost it becomes PIERRE's weekly worklist, and when an article lands the ledger replays the stored real questions and shows the confidence delta. The golden set stops being hand-picked and starts growing out of live traffic.

The promote-to-KB queue exists but nothing forces a root cause into it, so the loop currently turns without carrying anything. The gate is what loads it. Note the distinction it needs: a partner marking their work done is not the same event as the case closing.

E

Owners and review dates with teeth

ownerId · verifiedAt · nextReviewAt

The half of the metadata contract that's missing today. On its own a review date is decoration, so it has to be enforced: overdue content is automatically downgraded — a firmware article unverified for six months drops to MEDIUM trust, one unverified for a year is quarantined.

WHAT'S ALREADY THERE

TrustLevel and SourceStatus (active / deprecated / quarantined) already exist and are already honoured by retrieval. This is a scheduled job and four columns, not a subsystem. The effect is that the knowledge base cannot rot quietly, and the owner has a reason to touch it.

F

Escalation, callback and Engineering Check

EscalationPolicy · entry gate · output contract

The survey's clearest structural request: a real L1 → L2 → developer path with committed response times, and a route for out-of-hours. Per tenant and severity: target tier, response target, after-hours path. The assistant's job is to classify severity, attach the diagnostic context it already collected, and tell the installer what happens next and by when.

The operational concept sharpens this into two encodable gates. A callback is not the first support step — it unlocks on stated conditions (no safe step left, checks done and the fault persists, Portal data self-contradictory, several units affected, suspected global fault, PIERRE-side access needed, partner standing on site). And Engineering Check becomes a real state with an entry gate — version, reproduction steps, checks already run, users affected, business impact — and, more valuable, a mandatory output: cause or current status, temporary workaround, option A or B, owner, next step, committed date, fixing version.

WHAT'S ALREADY THERE

Worth saying plainly: this is an organisational commitment, not a software feature. The software's contribution is to make it visible and measurable — time to first human response, per tier, in the admin console. Half the frustration in the survey isn't the slow answer; it's not knowing whether an answer is coming at all.

The mandatory output is the part that closes the circuit: an Engineering Check that names a fixing version writes straight back into the known-issue registry. Two small rules ride along — the assistant never hands out a personal support phone number, and it asks whether the partner is on site, because that changes the queue position.

G

Site kit and field mode

pre-departure checklist · offline PWA · voice

The strongest single idea in the operational concept, and it inverts ours. We had a “before you leave site” checklist. The better one runs before the partner sets off: what will I need at this site? — generated from the project, the device and the reported symptom. Laptop and drivers, USB-to-RS485 adapter, LAN cable, multimeter, wiring diagrams, the Modbus register map, firmware and config files, a spare unit, access credentials, the site contact, whatever PIERRE-side remote access is required.

Around it, the rest of field mode: an offline cache scoped to the products and procedures in this installer's projects rather than the whole knowledge base, voice input, and one-tap diagnostic collection that pulls live device status into the case.

WHAT'S ALREADY THERE

Half of a repeat visit isn't caused by the fault surviving — it's caused by the adapter or the register map being at the office. Attacking that before departure is far cheaper than anything we can do afterwards, and every input it needs is already in the Portal and B2B.

The transcribe module and the PWA service worker are in place; voice also means the partner dictates the site report and gets a written record back rather than typing a ticket. The metric that matters is repeat site visits avoided — one question at ticket close. It is the number that justifies the subscription, because the partner can price it.

H

Partner competence profile

answer depth, never answer safety

From the operational concept: the assistant should work out whether it is talking to a seasoned integrator or someone doing this for the first time. Compass already has four answer modes — quick, step-by-step, diagnostic, deep engineer — but the mode is chosen per question, never per partner. An experienced integrator wants “check register 40001”; a beginner needs the walkthrough, and giving each of them the other one wastes the same amount of time.

WHAT'S ALREADY THERE

Derivable from what we already store: ticket history, which steps this partner completes unaided, how often escalation follows. It rides on the existing mode selector rather than adding a parallel path.

One boundary is not negotiable, and it should be written down before the feature is: competence adjusts depth and tone only — never the confidence thresholds, never the citation requirement. A senior partner gets a shorter answer, not a less-evidenced one.

In schema terms this is roughly seven new models and five new tools on top of the current 22 and 11, plus two admin surfaces — the gap ledger and the known-issue register. Nothing here requires re-architecting what exists.

SECOND SOURCE

Merged in: the operational concept

A second internal document, written from the support side, described how the service should run day to day. Where our plan is strong on data model and safety logic it is quiet on process; where it is strong on the partner's actual working day, ours was thin. Eight of its elements are folded into the proposal above.

G

The pre-departure site kit

Generated from project, device and symptom. Attacks the repeat-visit cost before the drive rather than after it — the highest-leverage idea in either document.

A

Feature → minimum-version matrix

Turns the version axis from a retrieval filter into something that answers a question partners ask constantly: why doesn't this work on my system?

A

Update-risk classification

Routine, needs on-site presence, needs PIERRE approval — plus the hard rule that the assistant never initiates an update or edits a configuration.

D

Closure gate and external-cause taxonomy

No close without a recorded cause. The external-cause list quietly measures what share of reported faults were never ours.

F

Engineering Check entry gate and output contract

More concrete than our tier model, and the mandatory output — cause, workaround, owner, date, fixing version — is what writes back into the known-issue register.

F

Callback preconditions

Ten stated conditions under which a phone callback is warranted. Encodable as a tool guard rather than left to judgement.

H

Partner competence profiling

Answer depth chosen per partner, not only per question. New — we had no equivalent.

C

Modbus, camera and network parameter sets

Baud rate, parity, slave ID, register map, RTU vs TCP, ports, VLAN, DHCP vs static. In effect a specification for the first batch of diagnostic procedures, from expert knowledge — useful as a seed while the gap ledger fills.

WHERE IT NEEDS COMPLETING

No account of uncertainty

The concept is written as though the assistant always has an answer. Nothing on what happens when it doesn't: no citation requirement for electrical claims, no refusal to infer wiring from a photo, no confidence at all. That is the property Compass is most careful about, and it has to survive the merge — a capability list is not a system until it says how it fails.

No gap measurement

Same blind spot as the partner survey: it assumes the knowledge is there. The process diagram is sound and would run over an empty knowledge base without complaining.

Commerce pulled into support

Orders, order status and missing items are a different permission model and a different freshness requirement from technical support. Worth doing — in its own phase, not folded into this one.

The tiers are unstaffed on paper

Callback and Engineering Check both assume named people and a response commitment. The concept specifies the process precisely and is silent on who answers and how fast — the same organisational gap the survey pointed at.

The two documents don't compete. Read together they cover both halves of the same system — one says what the machine is made of, the other what a Tuesday looks like for the person using it.

OUT OF SCOPE — DELIBERATELY

What we wouldn't build

Saying this out loud matters as much as the proposal: each of these is a plausible-sounding request that would cost real money and make the system worse.

A second product master

B2B is the source of truth and the spec keeps it that way. This is worth correcting explicitly, because “digitise the product catalogue” reads as a green light to start one.

Another support portal UI

Tickets live in the Operation Portal. Two places to look is worse than one place with gaps.

Four hand-maintained languages

Author once, translate at answer time. Maintaining HU / SR / HR / EN copies of every procedure is how knowledge bases die. The exception is safety-critical text — forbidden connections, warnings — reviewed once per language and then frozen as fixed strings, never translated on the fly.

Authoring ahead of the ledger

Ingest all 300 pages and 200 videos — that's cheap and automated. But nobody writes new material until the gap ledger says what's worth writing. Otherwise 80% of the effort goes to content no one ever asks for.
SEQUENCE

Three phases

Ordered so that everything content-dependent comes last, and the first deliverable arrives before PIERRE has written a single new page.

PHASEEFFORTCONTENTBLOCKED ON PIERRE?
Phase 0~2 weeksVersion axis + feature matrix + update-risk flags · gap ledger · owner metadataNo — pure plumbing
Phase 1~4–5 weeksKnown-issue registry · proactive alerts · escalation + Engineering Check · site kitPartly
Phase 2~4–6 weeksDiagnostic procedures · draft-from-ticket generator · competence profile · field modeYes

Running alongside from day one: live on internal support only. Their questions fill the ledger, and the ledger decides what Phase 2 is actually about.

NEEDS A DECISION

Three open questions

None of these block starting. All three block finishing.

Does B2B track firmware releases?
Decides whether the version axis is a connector or a new master. Blocks Phase 0 estimation, nothing else.
Who owns knowledge at PIERRE?
A named person per product area. Without owners the review dates have nothing to point at, and the ledger produces a list nobody is accountable for.
What response times will PIERRE commit to?
Per tier, including out of hours. The system can enforce and measure a commitment — it cannot invent one, and publishing an SLA nobody staffs makes the gap more visible, not smaller.
IN ONE SENTENCE

The knowledge base and the assistant don't come one after the other — the assistant is the only instrument that reliably says which knowledge is worth digitising, and without it the list of twenty-odd problems stays a guess.