● B2B offer — Purchasing & supply
Purchasing strategy: your requirements projected, your suppliers compared
Buying well means knowing what will be needed, when, and on what terms. Your agent, connected to your ERP, projects the requirements from your consumption and your forecasts, matches the terms of your approved suppliers and proposes several costed supply plans. Hosted in France: your negotiated terms and your volumes stay with you. The purchasing department chooses the plan and commits the orders.
Hosted in France
Negotiated terms protected
GDPR & AI Act: governed deployment
Human oversight
Updated on
Deployed in a few weeks
Purchasing strategy · hosted in France
What supply plan for the coming quarter?
Requirements projected by family, from your consumption and the forecasts entered in the ERP.
Three costed plans: each with its volumes, its approved suppliers and the applicable terms.
The projection assumptions are shown with each plan.
🔗 Sourced · ERP consumption and supplier terms
Three costed plans: each with its volumes, its approved suppliers and the applicable terms.
The projection assumptions are shown with each plan.
🔗 Sourced · ERP consumption and supplier terms
Which of the three is best?
Each optimises a different criterion: unit cost, security of supply, flexibility of volume.
The weight to give each criterion depends on your purchasing strategy — that call is yours, as is committing the orders.
✎ Support · criteria set out, the department's call
The weight to give each criterion depends on your purchasing strategy — that call is yours, as is committing the orders.
✎ Support · criteria set out, the department's call
Local inference · no data outside the EU
Plans hosted in France
Sovereign by designLocal inference or hosting in France
GDPR & AI Act: governed deploymentTraceability & human oversight
TurnkeyDesigned, installed and operated for you
The purchasing department decidesThe agent prepares, never rules
✦ In brief
A Blue Lemon Agent purchasing strategy agent, connected to your ERP, projects requirements from your consumption and your forecasts, matches the terms of your approved suppliers and proposes several costed plans, with the assumptions shown. Choosing the plan and committing the orders stay with the department. It runs on local inference or is hosted in France: your negotiated terms stay with you, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
100%
hosted in France in the target architecture
0
transfer outside the EU in the target architecture
6
purchasing uses ready to deploy
0
decision taken without human approval
These figures describe our offer, not results measured at a client. How large the gain is on your number of purchasing families and suppliers is confirmed by a pilot.
The context
What does an AI agent bring to your purchasing strategy?
A supply plan costed in advance opens the negotiation; built in a rush, it closes it.
! The issue
Steering purchasing means cross-checking projected requirements, supplier terms and strategic criteria. Building a single complete plan already takes time; comparing several takes more. The agent produces several, each optimising an explicit criterion, with its projection assumptions shown.
✓ Our answer
The purchasing department compares complete plans and decides the weight to give each criterion: cost, security of supply, flexibility. Committing an order remains its act. Local inference or an isolated resource hosted in France: your negotiated terms and your purchase volumes, decisive information in a negotiation, do not leave the company.
The decisive point
Your negotiated terms and your purchase volumes: sovereignty & compliance
Your negotiated terms and your purchase volumes bear directly on your margins. Here is how they are protected.
Local inference
The agent can run on a machine belonging to your organisation: no negotiated term and no purchase volume leaves the network.
Hosting in France
Otherwise, a dedicated and isolated resource hosted in France, under French law — your requirements, your suppliers and your forecasts: processing and access within the European Union targeted by the architecture.
Reduced extraterritorial exposure
For your negotiated terms and your purchase volumes, the architecture aims to reduce exposure to the Cloud Act and FISA 702; being located in France or in the European Union does not, on its own, guarantee immunity.
Isolated resource
No pooling: an environment strictly dedicated to your company and its supplier panel.
Projection assumptions shown
Every plan states the consumption, forecasts and terms used; encryption, role-based access and logging of the plans produced.
AI Act: governed deployment
The agent is strictly in support; no order is committed and no supplier is chosen automatically; traceability and human oversight from end to end.
What depends on the architecture chosen These points are not general guarantees: they are settled deployment by deployment, in the quotation.
- The applicable location is that of the architecture set out in the quotation and verified before commissioning.
- Local execution is announced only for the configuration explicitly described and accepted in the quotation.
- The applicable isolation depends on the deployment mode set out in the quotation; no dedicated isolation is presumed.
- Roles and permissions are configured and accepted for the identities and systems actually connected.
- The events logged, their content, their retention period and who may access them are defined for the deployment chosen.
For terms under a non-disclosure agreement and negotiations in progress, SecNumCloud and reinforced hosting are options depending on your requirements. A single architecture is designed to answer both the GDPR and extraterritorial exposure. Designed for deployment in line with the GDPR and the AI Act, after the processing, roles and context-specific risks have been assessed.
Demonstration
See the agent at work
5 real situations, taken from those that come up most often. Pick one: the exchange unfolds as it would in your organisation.
A scripted demonstration. These exchanges show how the agent behaves — its sources, its refusals, what it leaves to your teams. Nothing is sent from this page, no model is queried here, and the matters named are fictional. That is precisely what we promise your data.
The behaviours shown here — monitoring, automation rules, routing and reminders — are configured with you during deployment, from your tools, your rules and your thresholds.
The architecture points named in these exchanges — location, local execution, isolation, encryption, role-based access, logging — are not a guarantee attached to the demonstration: they are those of the architecture set out in your quotation, and verified before commissioning.
The company in this demonstration
Fictional companyVercors Thermique — manufacturer of heat exchangers and industrial heating systems
- Sector
- Industrial thermal equipment manufacturing, sold direct and through installers
- Headcount
- 210 staff, including 6 in procurement: 1 head of procurement, 3 category buyers, 2 supply planners
- Market served
- French and German food-processing and chemical manufacturers
- Purchasing volume
- €18M of purchases a year, 9 purchasing categories, 143 referenced suppliers, 4,200 order lines a year
- Tools in place
- ERP live for 9 years, framework contracts and price lists in the document store, shared procurement mailbox — the agent plugs into them, nothing is replaced
- Who decides
- The head of procurement picks the plan and commits the orders; category buyers negotiate; the industrial director sets cover levels
- Room for improvement
- A quarterly supply plan takes 11 days of work and exists in a single version; projecting needs takes 55 % of that effort and costing 50 %
Vercors Thermique buys well but always under pressure: the quarterly plan arrives once the quarter has started, and procurement has never had time to compare two of them. The agent runs on local inference on a machine at the plant and plugs read-only into the ERP, the contract store and the procurement mailbox: it projects, costs and drafts the orders; the head of procurement picks the plan and signs the mandate. The exchanges below cover one quarter, from the first projection to the review.
This company, its figures and the exchanges that follow were invented for the demonstration. They illustrate a common situation; they describe no real client.
Procurement planning · hosted in France
I read three years of your consumption last night — 4,200 order lines a year, 9 categories — and your supply plan for the quarter is projected, assumptions on show.
A supply plan is the costed calendar of what you buy: what quantity, from which supplier, on which date.
The gap I measured, and it is the one that carries the money: 924 of last year's lines — 22 % — went out as emergency replenishment, paid on average 18 % above your contract price. Across the year, €148,000 of price gap, of which 91 % sits in three categories only: high-temperature seals, stainless tubing, control boards. These are not accidents, they are three categories whose needs nobody has time to project.
What I am bringing you this morning: the 9 categories projected across the quarter, each with its central quantity, its low and high range, and the reference periods used. Total projected: €4.47M of demand, at ± 6 %.
The time this moves: projecting needs used to take 55 % of the exercise; it takes 8 %. Eleven days of work become one day, and that day goes into negotiating.
What I propose: you review the assumptions category by category, change any of them in one word, and I move straight on to reconciling your supplier terms. needs-projection_9-categories-quarter.pdf€4.47M projected at ± 6 %, assumptions per category
⛓ Sourced · ERP (3 years of consumption, 12,600 order lines), saved forecasts, contract prices
A supply plan is the costed calendar of what you buy: what quantity, from which supplier, on which date.
The gap I measured, and it is the one that carries the money: 924 of last year's lines — 22 % — went out as emergency replenishment, paid on average 18 % above your contract price. Across the year, €148,000 of price gap, of which 91 % sits in three categories only: high-temperature seals, stainless tubing, control boards. These are not accidents, they are three categories whose needs nobody has time to project.
What I am bringing you this morning: the 9 categories projected across the quarter, each with its central quantity, its low and high range, and the reference periods used. Total projected: €4.47M of demand, at ± 6 %.
The time this moves: projecting needs used to take 55 % of the exercise; it takes 8 %. Eleven days of work become one day, and that day goes into negotiating.
What I propose: you review the assumptions category by category, change any of them in one word, and I move straight on to reconciling your supplier terms. needs-projection_9-categories-quarter.pdf€4.47M projected at ± 6 %, assumptions per category
⛓ Sourced · ERP (3 years of consumption, 12,600 order lines), saved forecasts, contract prices
Our sales forecasts are in the ERP, but everyone knows they run optimistic.
They run 9.4 % high on average, and I measured it across twelve quarters rather than assuming it. The correction is already applied, category by category.
What your own history says: for each of the last 12 quarters I compared the forecast entered with the consumption actually recorded.
· Average gap: +9.4 % of forecast above actual.
· Largest gap: +21 % on control boards, and it is stable quarter after quarter — that is a bias, not noise.
· Two categories land within 3 %: flanges and fasteners. They are forecast by the workshop, not by sales. The bias does not come from your people, it comes from where the forecast is entered.
What I did with it: I apply to each category its own recalibration coefficient, not a blanket average — an average would have cut fasteners by 9 % for no reason. The recalibrated projection moves from €4.47M to €4.22M of central demand, and the range tightens from ± 6 % to ± 4.5 %.
What that is worth: across the twelve past quarters, this recalibration would have avoided an average of €96,000 of tied-up stock per quarter — goods bought too early, sitting still.
What I propose: I republish the coefficients every quarter with the measurement behind them. You watch them move, and you override one without touching the other eight. projection-assumptions_recalibration-by-category.pdf+9.4 % bias measured, 9 coefficients, ± 4.5 %
⛓ Sourced · ERP, 12 quarters of entered forecasts set against recorded consumption
What your own history says: for each of the last 12 quarters I compared the forecast entered with the consumption actually recorded.
· Average gap: +9.4 % of forecast above actual.
· Largest gap: +21 % on control boards, and it is stable quarter after quarter — that is a bias, not noise.
· Two categories land within 3 %: flanges and fasteners. They are forecast by the workshop, not by sales. The bias does not come from your people, it comes from where the forecast is entered.
What I did with it: I apply to each category its own recalibration coefficient, not a blanket average — an average would have cut fasteners by 9 % for no reason. The recalibrated projection moves from €4.47M to €4.22M of central demand, and the range tightens from ± 6 % to ± 4.5 %.
What that is worth: across the twelve past quarters, this recalibration would have avoided an average of €96,000 of tied-up stock per quarter — goods bought too early, sitting still.
What I propose: I republish the coefficients every quarter with the measurement behind them. You watch them move, and you override one without touching the other eight. projection-assumptions_recalibration-by-category.pdf+9.4 % bias measured, 9 coefficients, ± 4.5 %
⛓ Sourced · ERP, 12 quarters of entered forecasts set against recorded consumption
What about new part numbers? We are launching two ranges this year with no history at all.
47 new part numbers, and they are already projected: 44 by pairing with a close existing part, 3 through a wide range that I tighten within three weeks.
How I pair them: I match the bill of materials of the new part against those you already consume — material, diameter, pressure class, temperature range — and reuse the consumption profile of the closest one. Across the 44, the paired part shares at least four characteristics out of five.
What that gives, checked against the past: I replayed the method on the 61 part numbers you launched over the last three years. 52 would have been projected within 15 % of actual from the first quarter — that is the benchmark I give you to judge today's 47.
The 3 parts with no twin — the control boards of the connected range, which have no equivalent at your site: I start from a wide range, 400 to 900 units over the quarter, and the protocol that tightens it is already written: the first three weeks of real consumption, read every Monday, bring the range under ± 12 % by week four. To cover those three weeks without tying up cash, I propose a first order at 400 units, the bottom of the range, with a top-up option already negotiable with both referenced suppliers.
The quarter is projected end to end. Next: I run your 143 suppliers and their terms through the mill, and show you what you have already paid above your own contract.
How I pair them: I match the bill of materials of the new part against those you already consume — material, diameter, pressure class, temperature range — and reuse the consumption profile of the closest one. Across the 44, the paired part shares at least four characteristics out of five.
What that gives, checked against the past: I replayed the method on the 61 part numbers you launched over the last three years. 52 would have been projected within 15 % of actual from the first quarter — that is the benchmark I give you to judge today's 47.
The 3 parts with no twin — the control boards of the connected range, which have no equivalent at your site: I start from a wide range, 400 to 900 units over the quarter, and the protocol that tightens it is already written: the first three weeks of real consumption, read every Monday, bring the range under ± 12 % by week four. To cover those three weeks without tying up cash, I propose a first order at 400 units, the bottom of the range, with a top-up option already negotiable with both referenced suppliers.
The quarter is projected end to end. Next: I run your 143 suppliers and their terms through the mill, and show you what you have already paid above your own contract.
We have framework contracts everywhere, amendments, price lists. Nobody really knows what applies.
It is all reconciled, and you have €61,400 to recover over the last twelve months. The 314 lines involved are listed with their order numbers.
A framework contract is the agreement that fixes in advance the terms applying to every order in a period — prices, tiers, lead times.
What I read in your document store: 118 framework contracts, 96 price lists and 61 amendments, across 143 referenced suppliers. For every supplier I rebuilt the term actually in force on the day of each order.
What the reconciliation shows:
· 314 lines invoiced at list price while the rebate tier had been reached. A rebate tier is the volume threshold above which a lower price applies as of right. €61,400, of which €38,200 with a single supplier — the tier triggers on cumulative annual volume, and nobody was cumulating.
· 27 lines paid below contract price, in your favour: €4,100. I flag those too, because a reconciliation that only reports what suits you does not survive three minutes in front of a supplier.
· 9 suppliers out of 143 carry 82 % of the gap. You do not have 143 conversations to hold, you have nine.
The time this moves: reconciling terms used to take 35 % of the exercise; it takes 6 %.
What I propose: the 9 recovery letters are written, each with the table of lines involved, the contract clause that grounds the claim and the amount. You sign them, they go out within the hour. terms-reconciled_314-lines.pdf€61,400 earned, 9 suppliers, 9 letters ready
⛓ Sourced · document store (118 framework contracts, 96 price lists, 61 amendments), ERP (12 months of invoiced lines)
A framework contract is the agreement that fixes in advance the terms applying to every order in a period — prices, tiers, lead times.
What I read in your document store: 118 framework contracts, 96 price lists and 61 amendments, across 143 referenced suppliers. For every supplier I rebuilt the term actually in force on the day of each order.
What the reconciliation shows:
· 314 lines invoiced at list price while the rebate tier had been reached. A rebate tier is the volume threshold above which a lower price applies as of right. €61,400, of which €38,200 with a single supplier — the tier triggers on cumulative annual volume, and nobody was cumulating.
· 27 lines paid below contract price, in your favour: €4,100. I flag those too, because a reconciliation that only reports what suits you does not survive three minutes in front of a supplier.
· 9 suppliers out of 143 carry 82 % of the gap. You do not have 143 conversations to hold, you have nine.
The time this moves: reconciling terms used to take 35 % of the exercise; it takes 6 %.
What I propose: the 9 recovery letters are written, each with the table of lines involved, the contract clause that grounds the claim and the amount. You sign them, they go out within the hour. terms-reconciled_314-lines.pdf€61,400 earned, 9 suppliers, 9 letters ready
⛓ Sourced · document store (118 framework contracts, 96 price lists, 61 amendments), ERP (12 months of invoiced lines)
Where exactly do the tiers come from? If we write to a supplier, we need to be solid.
From your own documents, and every amount points to a dated page you can open in front of the supplier.
What I read in a contract, and set out in black and white: the unit price per part, the tier and its base — annual, quarterly or per order volume, which is where it all plays out —, the validity period, the indexation clause and its index, payment terms, late penalties.
How every line stays traceable: « tier 2 reached on 14/09, framework contract of 03/02, article 4.2, page 7 ». You open the page, it says what I wrote. No amount rests on a reading you cannot redo in ten seconds.
Where those documents come from, and this is what makes the exercise possible: only from your systems — your document store, your ERP, your procurement mailbox. Your negotiated terms never leave the company to be read, which is exactly what your buyers want to hear: what you pay at Sorenca stays unknown to everyone, starting with Sorenca.
What the reading turned up on top: 23 of your 96 price lists expire within 90 days, including 7 on categories in this quarter's plan. Without renewal those 7 categories revert to list price: +7.4 % on €1.1M, that is €81,000 over the quarter. The 23 renewal requests are written, with scope and annual volumes attached. Sign seven today and the subject disappears.
What I read in a contract, and set out in black and white: the unit price per part, the tier and its base — annual, quarterly or per order volume, which is where it all plays out —, the validity period, the indexation clause and its index, payment terms, late penalties.
How every line stays traceable: « tier 2 reached on 14/09, framework contract of 03/02, article 4.2, page 7 ». You open the page, it says what I wrote. No amount rests on a reading you cannot redo in ten seconds.
Where those documents come from, and this is what makes the exercise possible: only from your systems — your document store, your ERP, your procurement mailbox. Your negotiated terms never leave the company to be read, which is exactly what your buyers want to hear: what you pay at Sorenca stays unknown to everyone, starting with Sorenca.
What the reading turned up on top: 23 of your 96 price lists expire within 90 days, including 7 on categories in this quarter's plan. Without renewal those 7 categories revert to list price: +7.4 % on €1.1M, that is €81,000 over the quarter. The 23 renewal requests are written, with scope and annual volumes attached. Sign seven today and the subject disappears.
And when a supplier emails us a new price list, do you see it go past?
Yes, and I found nine that nobody had carried into the ERP. Two are price cuts: €12,300 you already negotiated and are not collecting.
What I watch: the shared procurement mailbox, read-only. An attachment carrying a price list is recognised by its content, not its file name — Sorenca's was called prices_2026_v3_final.pdf.
The nine lists received in six months:
· 2 cuts never applied: €12,300 on the coming quarter.
· 5 announced increases, of which 3 outside the indexation clause of their framework contract. For those three I have drafted the reply that recalls the contractual index and period: they represent €34,000 the contract does not oblige you to pay.
· 2 lists identical to those already on file, filed with no action.
What I propose, and this is what makes the gain permanent: a written price-writing mandate into the ERP. A mandate is a written, capped, dated authorisation, withdrawable in one word: it states exactly what I may execute alone.
· What it covers: updating a price in the ERP when it moves down and matches a price list signed by the supplier.
· What it caps: the 9 referenced categories, and nothing else.
· Its term: until 30/06, renewed in one word or withdrawn in one word.
· Its trail: every write is logged with the source list and its date of receipt.
Sign it and both cuts are in the ERP within four minutes, before your next order. Increases, on the other hand, stop and come back to your desk with the indexation clause and the costed comparison of the two other referenced suppliers: the negotiation is yours, and you open it armed. price-watch_9-lists-received.pdf€12,300 of earned cuts, €34,000 of out-of-clause increases
⛓ Sourced · procurement mailbox (6 months), 9 price lists received, framework contracts and indexation clauses
What I watch: the shared procurement mailbox, read-only. An attachment carrying a price list is recognised by its content, not its file name — Sorenca's was called prices_2026_v3_final.pdf.
The nine lists received in six months:
· 2 cuts never applied: €12,300 on the coming quarter.
· 5 announced increases, of which 3 outside the indexation clause of their framework contract. For those three I have drafted the reply that recalls the contractual index and period: they represent €34,000 the contract does not oblige you to pay.
· 2 lists identical to those already on file, filed with no action.
What I propose, and this is what makes the gain permanent: a written price-writing mandate into the ERP. A mandate is a written, capped, dated authorisation, withdrawable in one word: it states exactly what I may execute alone.
· What it covers: updating a price in the ERP when it moves down and matches a price list signed by the supplier.
· What it caps: the 9 referenced categories, and nothing else.
· Its term: until 30/06, renewed in one word or withdrawn in one word.
· Its trail: every write is logged with the source list and its date of receipt.
Sign it and both cuts are in the ERP within four minutes, before your next order. Increases, on the other hand, stop and come back to your desk with the indexation clause and the costed comparison of the two other referenced suppliers: the negotiation is yours, and you open it armed. price-watch_9-lists-received.pdf€12,300 of earned cuts, €34,000 of out-of-clause increases
⛓ Sourced · procurement mailbox (6 months), 9 price lists received, framework contracts and indexation clauses
Give me the plan for the quarter. And I want to understand what I am choosing.
Three complete plans, all built on the same recalibrated needs and the same verified terms. What changes between them is the criterion I optimise — and I tell you what each one concedes.
· Plan A — lowest unit cost: €4.42M. It concentrates volume to reach the top tiers. What it concedes: 63 % of volume with 2 suppliers, and 4 weeks of cover only. Cover is the number of weeks of consumption your stock and open orders let you hold.
· Plan B — security of supply: €4.58M. Three sources per critical category, 7 weeks of cover, no supplier above 28 % of volume. What it concedes: €160,000 more than plan A, that is 3.6 %.
· Plan C — volume flexibility: €4.51M. Open orders with revisable quantities, monthly call-offs. What it concedes: €90,000 more than plan A, and it requires holding a call-off calendar — 2 hours of planner time a month, which I carry.
The figure that decides, and it is yours: last year's shortages cost 3 line stoppages, 11 hours, €47,000, all on single-sourced categories. The €160,000 gap between A and B buys off those €47,000 of risk and 3 extra weeks of cover — exactly the trade-off your industrial director has been asking for over two years, and you now have it costed on both sides.
The time this moves: costing a plan used to take 50 % of the exercise; it takes 9 % — and you compare three where you used to have one.
What I propose: tell me which one serves as the base, and I build the variant you want to test. supply-plans_three-scenarios.pdf€4.42 / 4.58 / 4.51M, assumptions and concessions
⛓ Sourced · recalibrated needs, verified terms, 3 years of shortage and line-stoppage history
· Plan A — lowest unit cost: €4.42M. It concentrates volume to reach the top tiers. What it concedes: 63 % of volume with 2 suppliers, and 4 weeks of cover only. Cover is the number of weeks of consumption your stock and open orders let you hold.
· Plan B — security of supply: €4.58M. Three sources per critical category, 7 weeks of cover, no supplier above 28 % of volume. What it concedes: €160,000 more than plan A, that is 3.6 %.
· Plan C — volume flexibility: €4.51M. Open orders with revisable quantities, monthly call-offs. What it concedes: €90,000 more than plan A, and it requires holding a call-off calendar — 2 hours of planner time a month, which I carry.
The figure that decides, and it is yours: last year's shortages cost 3 line stoppages, 11 hours, €47,000, all on single-sourced categories. The €160,000 gap between A and B buys off those €47,000 of risk and 3 extra weeks of cover — exactly the trade-off your industrial director has been asking for over two years, and you now have it costed on both sides.
The time this moves: costing a plan used to take 50 % of the exercise; it takes 9 % — and you compare three where you used to have one.
What I propose: tell me which one serves as the base, and I build the variant you want to test. supply-plans_three-scenarios.pdf€4.42 / 4.58 / 4.51M, assumptions and concessions
⛓ Sourced · recalibrated needs, verified terms, 3 years of shortage and line-stoppage history
Plan A appeals on price. But if one of those two suppliers drops out, what do we do?
I measured what those two suppliers are worth, and built a tightened plan A that keeps 87 % of the saving while halving the dependency.
What your 3 years of goods receipts say: the supplier service level — the share of orders delivered complete on the promised date — is 97.1 % at Sorenca and 89.4 % at Halbrand. Halbrand carries 29 % of plan A volume and 7 of your 11 hours of line stoppage last year.
The plan A′ I built: I move 11 % of Halbrand's volume to your third referenced source on that category, whose service level is 96.3 %.
· Cost: €4.44M, that is +€14,000 on plan A.
· Concentration brought from 63 % down to 41 % on the top two suppliers.
· Expected shortage divided by 2.4, computed on your own service levels — applied to last year's €47,000 of stoppages, that is €28,000 less risk for €14,000 of spend.
The next step I propose, and it pays more than the trade-off itself: Halbrand holds €2.9M of your annual volume and you account for a meaningful share of their book. A 89.4 % service level, documented line by line, is a negotiating lever, not a worry. I build the file and you open the conversation next week. plan-a-tightened_concentration-63-to-41.pdf+€14,000, shortage risk divided by 2.4
⛓ Sourced · ERP (3 years of receipts, promised versus actual dates), line-stoppage history
What your 3 years of goods receipts say: the supplier service level — the share of orders delivered complete on the promised date — is 97.1 % at Sorenca and 89.4 % at Halbrand. Halbrand carries 29 % of plan A volume and 7 of your 11 hours of line stoppage last year.
The plan A′ I built: I move 11 % of Halbrand's volume to your third referenced source on that category, whose service level is 96.3 %.
· Cost: €4.44M, that is +€14,000 on plan A.
· Concentration brought from 63 % down to 41 % on the top two suppliers.
· Expected shortage divided by 2.4, computed on your own service levels — applied to last year's €47,000 of stoppages, that is €28,000 less risk for €14,000 of spend.
The next step I propose, and it pays more than the trade-off itself: Halbrand holds €2.9M of your annual volume and you account for a meaningful share of their book. A 89.4 % service level, documented line by line, is a negotiating lever, not a worry. I build the file and you open the conversation next week. plan-a-tightened_concentration-63-to-41.pdf+€14,000, shortage risk divided by 2.4
⛓ Sourced · ERP (3 years of receipts, promised versus actual dates), line-stoppage history
Then build me that file. The Halbrand meeting is Thursday.
The file is ready, and it fits in four pages you can read on the train.
What it holds, in the order it gets said:
· Your real weight: €2.9M placed in 2025, +34 % in three years, across 6 categories. You are a growing account, and that is the opening line of the meeting.
· Their record: 89.4 % service level, 141 lines delivered late, average delay 6.2 days, and the 3 dated line stoppages that followed. Every figure points to its order numbers.
· The three levers, costed: a volume rebate at −3 % (€87,000 over the quarter) · a service-level commitment at 95 % with a contractual penalty · a consignment stock on the two critical parts, which hands you 3 weeks of cover with no cash tied up.
· Your walk-away point: the costed comparison of the two alternative referenced suppliers, ready to put on the table. At equal terms, switching would cost you €46,000 a year — that is your room to move, and it is real.
The simulation, so you know what each concession is worth in the room: −3 % on the €2.9M placed with Halbrand is €87,000 over the quarter, in plan A as in plan B. Plan B, if retained, would move from €4.58M to €4.49M — the gap with plan A would fall to €70,000, and security of supply would become almost free.
What I propose for Thursday: I send you the file tonight and I hold the thread during the meeting — you dictate a counter-offer, I return its costed effect on the plan in under a minute, while you are still sitting across from them. negotiation-file_halbrand.pdf3 costed levers, −3 % = €87,000 a quarter
⛓ Sourced · ERP (volumes, receipts, delays), framework contracts, comparison of referenced suppliers
What it holds, in the order it gets said:
· Your real weight: €2.9M placed in 2025, +34 % in three years, across 6 categories. You are a growing account, and that is the opening line of the meeting.
· Their record: 89.4 % service level, 141 lines delivered late, average delay 6.2 days, and the 3 dated line stoppages that followed. Every figure points to its order numbers.
· The three levers, costed: a volume rebate at −3 % (€87,000 over the quarter) · a service-level commitment at 95 % with a contractual penalty · a consignment stock on the two critical parts, which hands you 3 weeks of cover with no cash tied up.
· Your walk-away point: the costed comparison of the two alternative referenced suppliers, ready to put on the table. At equal terms, switching would cost you €46,000 a year — that is your room to move, and it is real.
The simulation, so you know what each concession is worth in the room: −3 % on the €2.9M placed with Halbrand is €87,000 over the quarter, in plan A as in plan B. Plan B, if retained, would move from €4.58M to €4.49M — the gap with plan A would fall to €70,000, and security of supply would become almost free.
What I propose for Thursday: I send you the file tonight and I hold the thread during the meeting — you dictate a counter-offer, I return its costed effect on the plan in under a minute, while you are still sitting across from them. negotiation-file_halbrand.pdf3 costed levers, −3 % = €87,000 a quarter
⛓ Sourced · ERP (volumes, receipts, delays), framework contracts, comparison of referenced suppliers
We are going with plan B. How long before it becomes actual orders?
Six minutes after your signature. The 218 orders for the quarter are already drafted in your ERP, complete and checked.
What is ready: 218 orders, 41 suppliers, €4.58M, each with its part number, quantity, verified contract price, need-by date, incoterm and accounting code. Every line carries the number of the framework contract that grounds its price.
The time it hands back: your two supply planners used to key those orders in 31 hours of work — close to a week for two people. That time goes back into chasing and supplier quality, not into a keyboard.
The mandate I propose, written, capped, dated:
· Scope: the 9 categories of the chosen plan, the 143 referenced suppliers, and nothing beyond.
· Overall cap: €4.58M for the quarter, the amount of the plan you have just chosen.
· Per-order cap: €25,000. Above that, the order comes to you for signature — 14 of the 218 are in that case, and they are already ranked by urgency.
· Price: no order is issued above the verified contract price.
· Term: until 30/06. Withdrawable in one word, at any moment, including from your phone.
· Trail: every issue is logged — who covered it, which plan, which contract, to the second.
You sign the mandate, the 204 covered orders go out in six minutes, and you keep the 14 large ones in front of you. Then I hand you the book alive: acknowledgements, promised dates, gaps. purchase-mandate_capped-dated.pdf€4.58M capped, €25,000 per line, withdrawable in one word
✎ Framework · chosen plan, framework contracts, mandate settings, issue log
What is ready: 218 orders, 41 suppliers, €4.58M, each with its part number, quantity, verified contract price, need-by date, incoterm and accounting code. Every line carries the number of the framework contract that grounds its price.
The time it hands back: your two supply planners used to key those orders in 31 hours of work — close to a week for two people. That time goes back into chasing and supplier quality, not into a keyboard.
The mandate I propose, written, capped, dated:
· Scope: the 9 categories of the chosen plan, the 143 referenced suppliers, and nothing beyond.
· Overall cap: €4.58M for the quarter, the amount of the plan you have just chosen.
· Per-order cap: €25,000. Above that, the order comes to you for signature — 14 of the 218 are in that case, and they are already ranked by urgency.
· Price: no order is issued above the verified contract price.
· Term: until 30/06. Withdrawable in one word, at any moment, including from your phone.
· Trail: every issue is logged — who covered it, which plan, which contract, to the second.
You sign the mandate, the 204 covered orders go out in six minutes, and you keep the 14 large ones in front of you. Then I hand you the book alive: acknowledgements, promised dates, gaps. purchase-mandate_capped-dated.pdf€4.58M capped, €25,000 per line, withdrawable in one word
✎ Framework · chosen plan, framework contracts, mandate settings, issue log
And if a price moved between the plan and the order?
I check it on every line at the second of issue, and it happened seven times this morning. Five earn you money, two are on your desk with their solution already costed.
The five cuts: a list received on Friday takes two high-temperature seal parts below the plan price. −€2,400 applied immediately, under the price mandate you signed last week. The plan updates itself, downwards.
The two increases: +4.1 % and +6.8 % on stainless tubing, above the 2 % tolerance threshold I propose you set as a rule. They were pulled from the batch and did not go out. For each one you have in front of you: the contract's indexation clause and its index, the comparison with the two other referenced suppliers on the same part, and the exact effect on the plan — +€9,100 if you accept, −€1,200 and 4 extra days of lead time if you switch. The choice takes thirty seconds because the work is done.
What I propose writing into the mandate: a cut applies on its own, an increase stops and comes back to you documented. That rule is asymmetric, and deliberately so: it always plays your way.
And the threshold is yours: at 2 %, 2 lines a quarter come back to you; at 5 %, a single one — I costed both settings across the last twelve months so you choose on figures.
The five cuts: a list received on Friday takes two high-temperature seal parts below the plan price. −€2,400 applied immediately, under the price mandate you signed last week. The plan updates itself, downwards.
The two increases: +4.1 % and +6.8 % on stainless tubing, above the 2 % tolerance threshold I propose you set as a rule. They were pulled from the batch and did not go out. For each one you have in front of you: the contract's indexation clause and its index, the comparison with the two other referenced suppliers on the same part, and the exact effect on the plan — +€9,100 if you accept, −€1,200 and 4 extra days of lead time if you switch. The choice takes thirty seconds because the work is done.
What I propose writing into the mandate: a cut applies on its own, an increase stops and comes back to you documented. That rule is asymmetric, and deliberately so: it always plays your way.
And the threshold is yours: at 2 %, 2 lines a quarter come back to you; at 5 %, a single one — I costed both settings across the last twelve months so you choose on figures.
The real issue is afterwards: we find out about delays at goods-in.
Not any more: I match every acknowledgement to its order, and two date slips are on the table today, each with two costed answers.
What I track, order by order: acknowledgement received or not, promised date against need-by date, dispatch notice, actual receipt. 218 orders tracked without a planner opening a spreadsheet.
This morning:
· 3 orders with no acknowledgement at D+5. Chased automatically — that gesture commits nothing, it asks for a date, it promises nothing. Two answered the same day.
· 2 promised dates slip, one of them on control boards, which feed the assembly line from the 14th to the 21st. The impact is costed: 3 possible days of stoppage, €12,800.
The two ways out, both ready:
· Move the assembly batch by 4 days: no extra cost, but 2 customer deliveries move to week 12. I identified the two customers and drafted the notice messages.
· Cover from your second referenced source: +€1,900, available within 48 h, zero days of stoppage and no customer delivery moved. Covering costs €1,900 and saves €12,800: that is the one I recommend, and the draft order is ready.
What I propose next: day-to-day thresholds and replenishment are steered finely by the stock management agent, to which I already hand my projections. On your three critical categories, the pair would take emergency orders below 40 a quarter — you had 231. delivery-tracking_2-slips.pdf€12,800 of stoppage avoided for €1,900
⛓ Sourced · acknowledgements, dispatch notices, assembly schedule, customer delivery book
What I track, order by order: acknowledgement received or not, promised date against need-by date, dispatch notice, actual receipt. 218 orders tracked without a planner opening a spreadsheet.
This morning:
· 3 orders with no acknowledgement at D+5. Chased automatically — that gesture commits nothing, it asks for a date, it promises nothing. Two answered the same day.
· 2 promised dates slip, one of them on control boards, which feed the assembly line from the 14th to the 21st. The impact is costed: 3 possible days of stoppage, €12,800.
The two ways out, both ready:
· Move the assembly batch by 4 days: no extra cost, but 2 customer deliveries move to week 12. I identified the two customers and drafted the notice messages.
· Cover from your second referenced source: +€1,900, available within 48 h, zero days of stoppage and no customer delivery moved. Covering costs €1,900 and saves €12,800: that is the one I recommend, and the draft order is ready.
What I propose next: day-to-day thresholds and replenishment are steered finely by the stock management agent, to which I already hand my projections. On your three critical categories, the pair would take emergency orders below 40 a quarter — you had 231. delivery-tracking_2-slips.pdf€12,800 of stoppage avoided for €1,900
⛓ Sourced · acknowledgements, dispatch notices, assembly schedule, customer delivery book
A full quarter. What did it earn us, precisely?
€213,000 over the quarter, and every euro points to a document you can open.
· €61,400 of earned rebates never applied, recovered through the 9 signed letters.
· €12,300 of price cuts received by email and never carried into the ERP.
· €87,000 obtained at Halbrand, −3 % on €2.9M.
· €24,000 of price gap avoided on emergencies: emergency lines fall from 231 to 79 over the quarter.
· €28,300 on the supplier mix, plan B against simply rolling the existing one.
And the time, which never shows on an invoice: the planning exercise moves from 11 days to 1 day. On your three measured tasks: needs projection 55 % → 8 %, costing a plan 50 % → 9 %, reconciling terms 35 % → 6 %. Your three category buyers held 4 more negotiations this quarter — and that is where the €87,000 was found.
The figure that does not flatter me, and what I did with it: of the 9 categories, 7 landed inside my range. Two fell outside: −14 % on flanges, +11 % on control boards. The cause is clear: those two categories do not follow your seasonality, they follow customer tenders I had no sight of — one contract won in February and one postponed.
The correction is made and measured: for six weeks I have been plugged into your live sales pipeline, with the probability your sales team attaches to each deal. Over those six weeks the gap on those two categories fell back to 4 %, and it moved under 6 % across all nine. A projection is judged on its range and on how much hindsight backs it: this one has six weeks of hindsight, and I will hand you the measurement again at the full quarter. quarter-review_213000-euros.pdf5 sources of gain, 2 categories recalibrated
⛓ Sourced · ERP, recovery letters, Halbrand agreement, order log, procurement timesheets
· €61,400 of earned rebates never applied, recovered through the 9 signed letters.
· €12,300 of price cuts received by email and never carried into the ERP.
· €87,000 obtained at Halbrand, −3 % on €2.9M.
· €24,000 of price gap avoided on emergencies: emergency lines fall from 231 to 79 over the quarter.
· €28,300 on the supplier mix, plan B against simply rolling the existing one.
And the time, which never shows on an invoice: the planning exercise moves from 11 days to 1 day. On your three measured tasks: needs projection 55 % → 8 %, costing a plan 50 % → 9 %, reconciling terms 35 % → 6 %. Your three category buyers held 4 more negotiations this quarter — and that is where the €87,000 was found.
The figure that does not flatter me, and what I did with it: of the 9 categories, 7 landed inside my range. Two fell outside: −14 % on flanges, +11 % on control boards. The cause is clear: those two categories do not follow your seasonality, they follow customer tenders I had no sight of — one contract won in February and one postponed.
The correction is made and measured: for six weeks I have been plugged into your live sales pipeline, with the probability your sales team attaches to each deal. Over those six weeks the gap on those two categories fell back to 4 %, and it moved under 6 % across all nine. A projection is judged on its range and on how much hindsight backs it: this one has six weeks of hindsight, and I will hand you the measurement again at the full quarter. quarter-review_213000-euros.pdf5 sources of gain, 2 categories recalibrated
⛓ Sourced · ERP, recovery letters, Halbrand agreement, order log, procurement timesheets
What exactly do you do without asking us?
Three gestures, and all three play your way. Everything else lives in a mandate you write and withdraw in one word.
· Apply a price cut received in writing from a referenced supplier. €12,300 this quarter. An increase, by contrast, stops and comes back to you with its indexation clause and the comparison of alternatives.
· Chase a missing acknowledgement at D+5. The message asks for a date, it commits nothing. 34 chasers this quarter, 29 answers within 24 h.
· Recalibrate my own projection coefficients when real consumption contradicts them, and tell you — that is what brought the flange gap from 14 % to 4 %.
What the mandate you signed covers: issuing the orders of the chosen plan, within scope, under both caps, until 30/06. 218 orders issued, 0 above cap, 0 outside scope — and the log shows it line by line.
What I propose widening, costed: raising the per-order cap from €25,000 to €40,000 would take orders needing your signature from 14 to 4 a quarter. You gain 3 hours of signing; you keep your hand on the four lines that actually weigh. You decide, and that mandate is changed in one sentence. mandate-framework_what-runs-on-its-own.pdf3 automatic gestures, 2 caps, 0 breaches
✎ Framework · issue log, mandate log, chaser register
· Apply a price cut received in writing from a referenced supplier. €12,300 this quarter. An increase, by contrast, stops and comes back to you with its indexation clause and the comparison of alternatives.
· Chase a missing acknowledgement at D+5. The message asks for a date, it commits nothing. 34 chasers this quarter, 29 answers within 24 h.
· Recalibrate my own projection coefficients when real consumption contradicts them, and tell you — that is what brought the flange gap from 14 % to 4 %.
What the mandate you signed covers: issuing the orders of the chosen plan, within scope, under both caps, until 30/06. 218 orders issued, 0 above cap, 0 outside scope — and the log shows it line by line.
What I propose widening, costed: raising the per-order cap from €25,000 to €40,000 would take orders needing your signature from 14 to 4 a quarter. You gain 3 hours of signing; you keep your hand on the four lines that actually weigh. You decide, and that mandate is changed in one sentence. mandate-framework_what-runs-on-its-own.pdf3 automatic gestures, 2 caps, 0 breaches
✎ Framework · issue log, mandate log, chaser register
One last thing: our negotiated terms, our volumes. Where do they go?
Nowhere. I run on local inference on a machine at the plant, and your contracts never leave your network to be read.
Local inference means the model computes on your machine: the text of a framework contract crosses no outside network to be analysed. If you would rather not host a machine, the other route is an isolated resource hosted in France, dedicated to your company — no pooling with another manufacturer, and far less with one of your suppliers.
Why this is a purchasing argument and not a technical one: your negotiated terms and your volumes are exactly what a supplier would pay to know. Knowing you are at €2.9M with Halbrand and that your second source is €46,000 dearer is worth more, in the room, than three hours of preparation. Both of those facts stay inside your walls, under French law, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity, and they train no model.
What it looks like day to day: role-based access — rights follow the job: a category buyer opens their categories, not other people's terms —, encryption in transit and at rest, and a full log: who consulted which term, who covered which order, to the second. That log answers your statutory auditor in one page.
And you keep control wherever you are: a web dashboard, and supervision from your phone — you approve an above-cap order or withdraw a mandate in one message.
The next step I propose: the 9 categories are covered; your 4 subcontracting categories are not yet, and they weigh €3.1M. I have already read their 22 contracts: the exercise pays better there than anywhere else, because nobody has ever reconciled their terms. Say the word and I hand you the projection on Monday. technical-framework_where-your-terms-live.pdfLocal inference, role-based access, processing in the EU targeted
✎ Framework · deployment architecture, role-based access matrix, consultation and issue logs
Local inference means the model computes on your machine: the text of a framework contract crosses no outside network to be analysed. If you would rather not host a machine, the other route is an isolated resource hosted in France, dedicated to your company — no pooling with another manufacturer, and far less with one of your suppliers.
Why this is a purchasing argument and not a technical one: your negotiated terms and your volumes are exactly what a supplier would pay to know. Knowing you are at €2.9M with Halbrand and that your second source is €46,000 dearer is worth more, in the room, than three hours of preparation. Both of those facts stay inside your walls, under French law, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity, and they train no model.
What it looks like day to day: role-based access — rights follow the job: a category buyer opens their categories, not other people's terms —, encryption in transit and at rest, and a full log: who consulted which term, who covered which order, to the second. That log answers your statutory auditor in one page.
And you keep control wherever you are: a web dashboard, and supervision from your phone — you approve an above-cap order or withdraw a mandate in one message.
The next step I propose: the 9 categories are covered; your 4 subcontracting categories are not yet, and they weigh €3.1M. I have already read their 22 contracts: the exercise pays better there than anywhere else, because nobody has ever reconciled their terms. Say the word and I hand you the projection on Monday. technical-framework_where-your-terms-live.pdfLocal inference, role-based access, processing in the EU targeted
✎ Framework · deployment architecture, role-based access matrix, consultation and issue logs
Local inference · no data outside the EU
Your case is not here? That is exactly what a 15-minute conversation is for. Book the free audit →
Use cases
What does the agent actually do?
One agent, several purchasing levers. All these uses work in support, subject to your approval.
Included in your agent
The 3 capabilities essential to this promise are included, at no extra cost.
From 474 € excl. VAT / month
Projecting requirements
Projects volumes by family from consumption and forecasts.
Comparing terms
Matches the terms of your approved suppliers for each plan.
Several costed plans
Proposes several plans, each optimising an explicit criterion.
Controls and safeguards
These 3 controls are built into the agent: they frame what it does, whatever plan you pick. They are not chosen and are not added to your order.
Human validation, exceptions and escalation
Status, safe closure and audit trail
Record validations, conflicts of interest, criteria and decisions
What the agent must be connected to
This connection is required for the agent to work. It concerns your information system and is scoped during the audit.
Link enquiry, order, goods receipt and invoice without double entry
Need to go further?
These agents handle a different business process, with their own owner and their own price. They are added to this one.
Stock management
For day-to-day thresholds and replenishment, a dedicated agent takes them on.
Stock management agent (alerts, restocking) from 474 € excl. VAT / month Stock management →Round planning
For organising deliveries, a dedicated logistics agent completes the picture.
Logistics / round planning agent from 543 € excl. VAT / month Round planning →
Does your need fall outside this?
Book the free audit →
Build your agent
In 15 minutes we identify the most relevant agent — without oversizing the project.
The gain
How many plans can a purchasing department compare?
By taking on the projection and the costing, the effort shifts towards negotiation and strategy. How large the gain is depends on your volume and remains to be confirmed by a pilot.
Projecting requirements
Costing a supply plan
Matching supplier terms
Indicative figures, not contractual, to be confirmed by a pilot on your number of purchasing families and suppliers. The weight given to each criterion and the commitment of orders are matters of purchasing strategy: those decisions belong to the department.
How it works
The stages of your AI agent project
1
Audit & scoping
15 minutes to target the use case with the best return.
2
Quote or direct sign-up
A catalogue offer is bought online; a specific need gets a costed quote.
3
Design
We design the agent and its guardrails.
4
Integration & testing
We connect your tools to the agent, which is itself hosted in France.
5
Rollout
Going live and training your team.
6
Operation
Continuous supervision and improvement.
Pricing
One package, one agent
A purchasing strategy agent (projection, terms, costed plans), installed and operated for you.
Agility
Setup + controlled subscription
5,470 € excl. VAT setup
then 474 € excl. VAT/month — you invest at installation and pay a reduced subscription. Ideal for keeping the cost under control over time.
- Installation, configuration and training for your teams
- Operation, human oversight, updates and support
- Sovereign hosting in France, a dedicated and isolated resource
The simplest
Serenity
All inclusive, no setup fee
779 € excl. VAT /month
all inclusive, immediate start. No upfront investment: a single subscription. Ideal for starting quickly and simply.
- Setup included (installation, configuration, training)
- Operation, human oversight, updates and support
- Sovereign hosting in France, managed end to end
100% Sovereign
On site, you own it
8,805 € excl. VAT setup
then 674 € excl. VAT/month · + hardware from 1,058 € (one-off purchase, in addition) — a sovereign computer installed on your premises, maintained remotely. Models run locally, your data returned at the end of the contract. 36-month commitment.
- Hardware installed on your premises (you own it)
- French / European AI models run locally
- Secure remote maintenance (Pro support included)
Not included in the packages: AI consumption (model tokens), re-invoiced at real cost with no margin, and tracked in real time in your client area. Maintenance and supervision subscription for an initial term of 12 months for the Agility package, 24 months for the Serenity package and 36 months for the 100% Sovereign package, renewable; support levels (SLA 72 h / 24 h / 4 h) optional. Bespoke development, additional integrations or exceptional volumes are quoted separately. Support Monday to Friday, 9am to 6pm. Prices exclude VAT.
AI model:
none of the AI models offered currently carries a fixed surcharge. When the selected model carries a cost, that cost is shown when you choose it, before you order, and re-invoiced at the cost incurred, with no mark-up; usage is billed at the publisher's price. Publishers' prices are published in US dollars: the amount re-invoiced is the amount in euros actually borne by Blue Lemon Agent on the publisher's invoice, at that invoice's exchange rate, with no commission or mark-up.
Included components and additional components Components included in the base offer: the Blue Lemon Agent software foundation, the AI models listed in the order journey, the standard channels (Microsoft Teams, Slack, WhatsApp Business, email, website chat, calendars, Microsoft 365 / Google Workspace, file storage, market VoIP telephony, professional social-media pages and accounts, Google Business Profile), hosting in France for the package chosen, backups, supervision, updates and support. If adapting the AI agent to your constraints, your needs or your requests requires other paid components — a third-party publisher's software licence, paid API access to one of your applications, hosting of health data, for which French law requires an HDS-certified host (art. L. 1111-8 of the French Public Health Code), SecNumCloud-qualified hosting, a speech synthesis service, particular hardware —, they are offered to you as an option or on quotation and re-invoiced at the cost incurred; nothing is committed without your written agreement. Where the artificial intelligence model you choose entails an additional cost, that cost is shown to you before you order and re-invoiced to you at the cost incurred, with no margin.
What to expect
Go-live
2 to 3 weeks
Agent designed, channels connected, team trained.
Steady state
4 to 7 weeks
After a few weeks of real use, once the agent's behaviour matches what you expect.
Indicative estimate, adjusted to the options you keep. It is not a delivery commitment.
Our commitment
Four guarantees that matter to your purchasing
Your negotiated terms stay with youLocal inference or an isolated resource hosted in France; no term and no volume entrusted to a third party, no data used to train a model.
Data in France, under French lawYour negotiated terms and your purchase volumes: minimisation and location in France, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
The purchasing department keeps the decisionThe agent produces costed supply plans, which can be checked and altered; no approval is automated.
Human oversight & traceabilityOn your number of purchasing families and suppliers: systematic logging and tracking, in line with the AI Act.
Going further
Related resources
Frequently asked questions
Your questions, our answers
Does the agent commit orders?
No. It proposes costed plans with their assumptions; choosing the plan and committing the orders belong to the purchasing department.
What are the projections based on?
On your past consumption and the forecasts entered in your ERP. The assumptions used are shown with each plan, so they can be discussed.
How are the suppliers compared?
On the terms of your approved panel, as they appear in your systems. The agent consults no external source.
How does this differ from stock management?
Stock management handles day-to-day thresholds; this strategy work projects requirements over several months and compares complete supply plans.
Are our terms protected?
Yes. The agent is hosted in France, on local inference or an isolated resource, with the deployment objective of processing and access operated within the European Union and an architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity. Your terms are not used to train a third-party model.
How long does it take to deploy this agent?
A few months as a rule, depending on the number of purchasing families and how deep the connection with your ERP goes, after a free audit then phases of design, integration and testing.
Related use cases
Other agents for the supply chain
Let's talk
Let's size up the potential in your purchasing
15 minutes to frame your purchasing families and your ERP — hosted in France, supervised, with no commitment.