Prepared for Odele Beauty leadership. Please enter the access code to continue.
The seam between NetSuite, Microsoft 365, True Commerce, and your 3PL portals is where ops time and AI leverage hide. Before adding another planning SaaS, see whether the operating layer can deliver it through dashboards and orchestration over data you already own. Specialists like Glimpse get called into where their depth earns it. Everywhere else, build the connective tissue instead of buying another login.
On the table right now: separate pre-AI subscriptions for demand, supply, and financial planning, plus AI-native specialists like Glimpse (chargebacks) and Emmy (procurement).
Generic Copilot retrieves. A point-solution AI runs one workflow. Neither owns the seam between MPS, NetSuite, the supplier mailbox, and the 3PL portal, where Odele's operating leverage hides.
A skill is the encoded answer to "how does Odele actually do this work?" MPS pulled from OneDrive. PO drafted in your NetSuite template. Supplier email in your voice. Reply tracked, escalated at 14 days. Audit logged. Nothing leaves before a human approves.
The platform compounds. Each skill costs less than the last. Memory accumulates per supplier, SKU, retailer. Tacit logic scattered across spreadsheets and Slack becomes executable.
Connectors fit Odele's existing tools. Runtime sits inside your M365 tenant. The approval queue is the permanent boundary between the model and any spend, send, or commitment.
Stage one is the hard part. Most platforms skip it and ship something that breaks the first time it touches a real PO.
Sit with the operator who runs the workflow today. Trace MPS-to-PO end-to-end. Capture the unwritten rules: which suppliers get which lead time, which SKUs are pre-approved for auto-draft, what Andrea would never send to a supplier without rephrasing.
Where the depth livesWrap the spec in a callable skill. Connect it to NetSuite, OneDrive, and the supplier mailbox. Implement eval logic, PO and email templates, audit hooks. Register the same skill as an MCP tool so Copilot 365 can call it.
Curated golden examples gate promotion. The MPS-to-PO skill has to match past POs Odele actually sent before it goes to the full team. Promotion is a measurable threshold, not a meeting.
The second skill (3PL rate quotes, monthly dashboard, supplier email triage) is cheaper than the first. Shared connectors, eval patterns, and memory layer. The Odele ops team becomes skill authors. The platform grows from inside, not from a vendor roadmap.
Every skill output lands as a draft with sources attached and an eval score. Approve, edit, or reject line by line. Same skills callable from Copilot 365 via MCP. Click a draft in the sidebar to inspect it, or load a skill from the input below.
Pick a skill from the sidebar, or click an existing draft to inspect.
3 pending drafts · click any row to inspect, approve, edit, or reject.
9 skills · Odele's evolving playbook codified into runnable, audit-logged agents. Click any card to see the spec.
142 entries · per-supplier, per-SKU, per-lane notes the platform recalls each cycle.
Every run, edit, approval, and rejection. Immutable, searchable, exported nightly.
An MPS spreadsheet on OneDrive, NetSuite inventory, True Commerce EDI, and the 3PL portals are enough to drive demand and supply planning views that hold up next to Anaplan or Kinaxis. The platform reads the data, runs the playbook, and renders it. No new contract, no separate login, no pre-AI SaaS suite. Where a specialist (Glimpse) genuinely earns its place, the platform calls into it.
| SKU | Channel | u/wk | Δ |
|---|---|---|---|
| Smoothing Shampoo 13oz | Target | 14,820 | +8% |
| Smoothing Conditioner 13oz | Target | 12,340 | +6% |
| Moisture Repair Shampoo 13oz | Ulta | 9,612 | +11% |
| Smoothing Shampoo Bulk 32oz | DTC | 8,201 | +19% |
| Curl Defining Conditioner 13oz | CVS | 3,140 | −14% |
| PO | Supplier | $ | Age |
|---|---|---|---|
| PO-2587 | Berlin Pkg | $22,400 | 11d |
| PO-2592 | Givaudan | $18,200 | 6d |
| PO-2604 | Aptar | $9,800 | 3d |
| SKU | Description | Channel | On-hand | Open POs | Forecast units · weeks W20 – W27 | Supplier | LT (d) | Wks cover | Reorder? | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| W20 | W21 | W22 | W23 | W24 | W25 | W26 | W27 | |||||||||
| SS-13-TGT | Smoothing Shampoo 13oz | Target | 128,400 | 42,000 | 14,800 | 14,900 | 15,200 | 15,400 | 15,600 | 15,800 | 16,000 | 16,200 | Coyne CoPack | 21 | 10.4 | — |
| SC-13-TGT | Smoothing Conditioner 13oz | Target | 96,200 | 36,000 | 12,200 | 12,400 | 12,600 | 12,800 | 13,000 | 13,200 | 13,400 | 13,600 | Coyne CoPack | 21 | 9.8 | — |
| MRS-13-ULT | Moisture Repair Shampoo 13oz | Ulta | 62,400 | 18,000 | 9,400 | 9,500 | 9,700 | 9,900 | 10,100 | 10,300 | 10,500 | 10,700 | Berlin Pkg | 35 | 7.2 | flag |
| SS-32-DTC | Smoothing Shampoo Bulk 32oz | DTC | 14,200 | 0 | 8,100 | 8,300 | 8,400 | 8,600 | 8,800 | 9,000 | 9,200 | 9,400 | Coyne CoPack | 21 | 1.7 | REPLAN |
| CDC-13-CVS | Curl Defining Conditioner 13oz | CVS | 9,800 | 0 | 3,100 | 3,140 | 3,180 | 3,220 | 3,260 | 3,300 | 3,340 | 3,380 | Berlin Pkg | 35 | 3.1 | REPLAN |
| MRC-13-ULT | Moisture Repair Conditioner 13oz | Ulta | 88,600 | 24,000 | 10,800 | 10,900 | 11,000 | 11,100 | 11,200 | 11,300 | 11,400 | 11,500 | Coyne CoPack | 21 | 10.2 | — |
| CS-13-TGT | Clarifying Shampoo 13oz | Target | 42,800 | 12,000 | 5,800 | 5,900 | 6,000 | 6,100 | 6,200 | 6,300 | 6,400 | 6,500 | Coyne CoPack | 21 | 8.9 | — |
| PS-13-CVS | Purple Shampoo 13oz | CVS | 38,400 | 10,000 | 5,200 | 5,300 | 5,400 | 5,500 | 5,600 | 5,700 | 5,800 | 5,900 | Coyne CoPack | 21 | 8.7 | — |
| ADS-6-DTC | Air Dry Styler 6oz | DTC | 19,200 | 8,000 | 2,800 | 2,900 | 3,000 | 3,100 | 3,200 | 3,300 | 3,400 | 3,500 | Berlin Pkg | 35 | 7.4 | flag |
| Totals (9 of 142 SKUs shown) | 500,000 | 150,000 | 72,200 | 73,240 | 74,480 | 75,720 | 76,960 | 78,200 | 79,440 | 80,680 | 2 replan | |||||
All figures above are mock data generated for the demo. The point: NetSuite + the MPS spreadsheet + EDI + 3PL portal data Odele already has are sufficient inputs. The platform reads them, runs the planning logic, and renders these views. Where genuine specialist depth matters (chargeback dispute analytics, complex tax provision modeling), the platform calls into Glimpse or the right specialist rather than rebuilding it.
Every card traces to a moment in our Mar 27, Apr 1, or May 4 calls. Phase placement is honest: lead with what we build well, defer to specialists where they earn it.
MPS into OneDrive → POs drafted in NetSuite → approval queue → supplier send → reply tracked, escalated at 14 days. Human approval before any send.
Reviewer queue UI where any agent output lands as a draft. Append-only audit log captures input, sources, model, reviewer, decision. Built once, reused everywhere.
Same skills exposed as MCP tools inside Copilot 365. Channels the shadow-AI experimentation Shanti flagged on Apr 1 through a tenant-aligned, audit-logged surface.
Lead times, packaging quirks, who-replies-fastest, which Target buyer prefers what cadence. Memory accumulates across runs so cycle 2 uses what cycle 1 learned.
Lane and volume in. Quotes out. Portal automation now, real APIs as carriers expose them. Side-by-side comparison with prior-cycle benchmarks.
Shanti's monthly Excel dashboard, populated automatically from NetSuite, EDI, and 3PL data in the layout the team trusts. Drill-down to source rows, threshold-delta flags. Read-only, fastest reusable win after MPS-to-PO.
EDI orders evaluated against rules + model judgment. Clean ones flow through; anomalies (qty, terms, SKU mismatch, chargeback risk) route to a human queue with the reason flagged. Addresses the line-by-line review pain Shanti named on both calls.
Curated prompts for the three tiers Shanti named (leadership, managers, frontline). Surfaced in Teams so Tier 3 doesn't learn prompt engineering. Answers Shanti's "people shouldn't have to become AI experts" principle.
Shanti's Mar 27 example: "How much did we spend on POs for volumizing shampoo last year?" answered across NetSuite, EDI, OneDrive without logging into anything. Built on P1+P2 connectors; P3 adds retrieval, role-based redaction, and the dashboard surface.
Extends P2 rate quotes into broader 3PL ops: inventory sync, shipment status, exception handling. One normalized surface in front of multiple carrier portals.
Tracks supplier replies against POs sent, parses confirmations and counter-offers, escalates anything quiet for 14 days into an Andrea queue. Closes the "did the supplier reply" loop.
Statistical baseline (seasonality, trend, retailer events) plus model judgment over POS and historical actuals. Output in the spreadsheet format Shanti's team already trusts, with confidence bands and "why this changed" notes per SKU.
Glimpse spans finance, sales, and supply chain for chargeback root-cause. Real depth, won't out-build it in 8 weeks. The platform calls into Glimpse and owns the integration, approval queue, and NetSuite/EDI links.
Kinaxis-class constraint solving across suppliers, lead times, capacity, sourcing. Hard math, years of tuning. At Odele's scale (one co-packer, modest supplier base) P3 is workable. Build-vs-buy decision worth revisiting if complexity grows.
Tax provision, deep scenario modeling, complex consolidations. Out of scope for the first six months. Platform feeds clean operating data to whatever financial tool Odele runs, doesn't try to be it.
Sources: Mar 27 discovery · Apr 1 architecture · May 4 SOW review. Phase order will likely shift after May 11.
MPS-to-PO is the anchor. The other two reuse the same connectors and approval pattern. Sequencing is Odele's call after May 11.
MPS into OneDrive → POs drafted in NetSuite → approval queue → supplier send → replies tracked, escalated at 14 days. Human approves before anything leaves.
Lane and volume in. Quotes out. Portal automation now, real APIs as carriers expose them. Side-by-side with prior-cycle benchmarks.
Shanti's Excel dashboard, populated from NetSuite, EDI, and 3PL data in the layout the team trusts. Drill-down to source rows, threshold-delta flags.
Same skills callable from a custom web UI and from Copilot 365 via MCP. Not two products to maintain. The team picks the surface that fits the moment, the platform stays single-source.
Skills act on Odele's actual data: NetSuite POs, OneDrive MPS, the EDI feed, the supplier mailbox. They do not free-generate values that should come from a system of record. Every output traces to its sources.
POs, supplier emails, and EDI sends land as drafts. Reviewer can approve, edit, or reject line by line. Nothing commits to a vendor without a human in the loop. Reviewer decisions train the next eval cycle.
Per-supplier, per-SKU, per-retailer context accumulates across runs. Lead times, packaging quirks, who-replies-fastest, which Target buyer prefers what cadence. The same skill called twice on the same supplier uses what it learned the first time.
Complex work calls multiple skills. The MPS-to-PO skill calls supplier-memory for lead times and supplier-mail for the send. The dashboard calls the same NetSuite reader the PO skill uses. Pipelines emerge from primitives.
The platform lives in a GitHub repo Odele can read, fork, or hand to another partner. No black box. No proprietary runtime. Continuity is structural, not a contract clause.
A wedge, not a cathedral. MPS-to-PO live and adopted in eight weeks teaches Odele more than six months of platform planning. The May 11 session decides whether MPS-to-PO is the right anchor or the team picks something tighter.
Eight weeks from kickoff to MPS-to-PO live with one real production cycle, the platform shell installed in Odele's M365 tenant, audit live, and weeks 5–8 already shipping the second skill.
Azure tenant, model access via MS infra, GitHub repo + CI, NetSuite service account scoped to the PO module, MCP server registered in Odele's Copilot tenant.
MPS ingest in Odele's format. PO drafting against the real NetSuite PO template. Approval queue UI. Supplier email send via the Odele service mailbox. One full real production cycle with full Odele oversight in week 4.
3PL rate quote pull or monthly dashboard, picked from the May 11 wishlist. Reuses the connectors and approval pattern. Pair sessions with SMEs to capture tacit logic. Role-based permissions hardened.
Review adoption and friction signals. Pick the next 2–3 workflows from the refined wishlist. Decide: continue the managed-service cadence, or pause to evaluate.
Forty-five minutes to walk the working slice and pick the first skill together. The eight-week sprint follows from the conversation.
Continue the Conversation