OMX innovation · Augmented Organisation · Straw man v0.1
Page 01 — The thesis

Augment the value stream.
Not the seat licence.

The default AI playbook is "buy everyone a Copilot licence and hope productivity lifts." That's a procurement decision, not a transformation. It puts tools on desks and leaves the work unchanged. OMX takes the opposite path — start with the value stream, find the friction and noise, then design a small, repeatable, governed agent that removes it. Same data. Same people. Less drag.

Anti-pattern
Seat-licence sprawl

Every staff member gets a Copilot / Claude / chat tool. Adoption is uneven. Output quality is uneven. Governance is unclear. Sensitive data leaks via opportunistic copy-paste. The bill compounds, the value doesn't.

The OMX pattern
Value-stream augmentation

Map the value stream. Pin the highest-friction step. Build a governed agent for that step — role-aware, business-aware, live-data-fed. Roll out the process, not the licence. Measure the friction removed, not the seats sold.

01 / 03
OMX innovation · Augmented Organisation · Straw man v0.1
Page 02 — The model

The pipeline. End to end.

Five moves, in order. Skip one and the system falls back to seat-licence sprawl.

01 — Identify
Map the value stream
Pick one stream end-to-end. Quote → cart → invoice. Lead → opportunity → close. Find every handoff, every queue, every re-key.
02 — Locate
Pin the friction
Where does the work wait? Where is the same data re-entered? Where do people email a screenshot? That's the friction. Name it.
03 — Encode
Repeatable process
The step a human does every time, the same way, with the same inputs and outputs. If it's not repeatable, it's not ready for an agent yet.
04 — Augment
Governed agent
An agent with role context, the OMX business knowledge baked in, and live data feeds. Auditable. Reversible. Scoped to one job.
05 — Measure
Friction removed
Cycle time before vs after. Error rate before vs after. Cost per transaction. Show the gain in the stream's own KPIs, not in licence counts.

What every agent must carry

Secure
Identity bound to the person and their role. No shared service accounts. Tenant-scoped. PII handled on the same rails as everything else.
Governed
Versioned prompts, change control, audit logs. Every output traceable to its inputs. Approval gates where the action is material.
Controlled
Bounded scope — does one job. Cannot do other people's jobs. Cannot escalate. Cannot exfiltrate. Fails closed, not open.
Role context
Knows who is asking, what they're allowed to see, what their team owns, and what the next step in the value stream looks like for them.
Business understanding
Knows OMX taxonomy, sector codes, customer hierarchy, replen rules, DC layout, contract floors. The institutional knowledge, encoded.
Live data feeds
Reads from Snowflake / Pronto / Salesforce in real time. Not a copy. Not a snapshot. The agent acts on the same numbers the human would see.
02 / 03
OMX innovation · Augmented Organisation · Straw man v0.1
Page 03 — Where we start

The OMX value streams. Each gets one agent.

Not theory. Real value streams already in flight or queued where the augmented-organisation pattern replaces friction we can name. Each is a candidate Phase 1 build. Same governance spine, different domain agent.

Stream 01 — Customer service
ISAM call handling
Friction: no CRM context on inbound call, manual lookup, customers re-identifying themselves.
Augment: Amazon Connect screen-pop with role-aware Salesforce summary + agent-assist next-best-action. Live data: SF account, recent invoices, open quotes.
Stream 02 — Quote turnaround
Self-service quoting
Friction: small / unmanaged customers ring a sales rep for items already in the catalog. Days, not minutes.
Augment: guided quoting agent reading contract pricing, MOQs, stock, lead time live. Customer sees their tier; rep sees override controls.
Stream 03 — Account research
Briefing Room · QBR pre-fill
Friction: sales managers spend half a day pulling the same numbers and slides before a quarterly review. Data exists; assembly is manual.
Augment: Company Research agent reading Snowflake + Salesforce + the 200 deep dives, producing a role-aware QBR brief on demand.
Stream 04 — Product onboarding
New SKU enrichment
Friction: new SKU lands with bare-minimum attributes. Web team scrambles, search misses, PDP looks thin, off-range margin leaks.
Augment: enrichment agent reading vendor spec + Icecat + competitor PDPs, drafting category attributes, classifying to L5 hierarchy, flagging human review where confidence drops.
Stream 05 — Digital comms · social
Lifecycle + GenAI social
Friction: campaigns built one-off; triggered comms hard-coded; social cadence depends on whoever has time. Brand voice drifts.
Augment: comms agent with brand tokens + campaign library + audience segments live from Snowflake; drafts trigger emails and social posts to a stewarded human queue.
Stream 06 — Web / internal SaaS build
Lens · Accelerate · WhoOnSite · Surface
Friction: packaged SaaS for every internal need = recurring licence + vendor roadmap captivity. Custom build feels slow.
Augment: build agents (already proven on Lens, Accelerate, WhoOnSite) — Snowflake-native thin web UI, OMX design language, shipped in weeks not quarters. Pattern compounds: each build cheapens the next.
Stream 07 — Product + category reviews
Range planning loop
Friction: review cadence is calendar-driven not signal-driven; merch + supply meet with spreadsheets prepared in different shapes. Decisions lag the data.
Augment: review agent with live sell-through, margin, stock-cover, supplier OTIF and competitor coverage — pre-builds the review pack and flags the SKUs that need a decision this cycle.
Stream 08 — Add yours
The list isn't closed
The pattern works wherever a repeatable step is gated by missing context, live data, or institutional knowledge.
Candidates we've heard so far: DC replen, contract renewal, supplier onboarding, RFP response. Bring the friction; we'll design the agent.
Straw man v0.1 — provoke a real conversation. This is a sketch, not a plan. Numbers will be sized once a stream is committed. The proposition is the shape: build agents that remove friction from value streams, not licences that sit on desks. Tear it apart and we'll rebuild.
03 / 03